BIOLOGICAL CHEMISTRY1. INTRODUCTIONTHE ramifications of biochemistry are now so extensive that it probablycomes as no surprise that one of the topics considered in this year’s Reporthas not been dealt with previously. A considerable amount of work iscurrently being undertaken on mineral metabolism with special reference toteeth; much of the information available is the result of fruitful co-oper-ation between scientist and clinician and is thus often published in journalsnot on the reading list of more chemically minded biochemists. A review ofthis work in Anlzual Reports will remedy this; furthermore, it is importantto have available the chemical and biochemical facts on such emotionallycharged problems as 9*Sr fall-out and fluoridisation of drinking water.The discovery that p-carotene was converted into vitamin A in the in-testinal mucosa of aninials and not in the liver was made during 1946-48.When it was iurther shown that the conversion in vivo proceeded easily, onefelt that the mechanism of the reaction would soon be revealed.Theproblem has, however, proved most complex and a stocktaking at the presenttime after 10 years’ endeavours seems appropriate.Progress in the study of muscular contraction has continued steadilyduring the past few years, but an important recent development in musclebiochemistry has been the attempt to elucidate the chemistry of relaxation.The characteristic cycle of normal muscle fufiction embraces both contractionand relaxation, and the chemical events associated with these are comple-mentary and equally important aspects of muscle biochemistry. Thisyear’s Report discusses modern aspects of both features in relation to earlierobservations.The discovery of new amino-acids is but one aspect of biochemistrywhich has been enormously aided by chromatography in all its variousaspects, and applications of these techniques to a study of higher plants haverevealed a number of interesting new compounds.Most of these have nowbeen well characterised chemically and the biosynthesis and metabolism .ofsome are now reasonably clear. However, the biochemistry of many ofthese new amino-acids remains to be investigated. ‘r. w. G.2. CHEMISTRY OF BONE AND TOOTH MINERALSTHERE has been a reawakening of interest in the chemistry of mineralisedtissues during the last decade.The skeleton is not merely a convenient framework upon which musclescan act and from which viscera depend. It is a vast storehouse which pro-vides minerals to the body in times of stress and is replenished in times ofplenty.The minerals of bone are distributed in an organic matrix, about70% by wt. of mature bone being inorganic. The tooth consists of twomajor tissues, enamel which is the most highly mincralised tissue in the bodyHARTLES: CHEMISTKY OF BONE AND TOOTH MINERALS. 323containing less than l:/b of organic matter, and dentine, which containsabout 7Sy0 of minerals. The third mineralised dental tissue, the cementum,covers the roots of teeth in a thin layer and is chemically similar to bone.Perhaps the most important physical characteristic of the mineralcrystals of bone and teeth is their smallness.Early workers,l basing theirestimates on X-ray-diffraction studies, deduced that the length of the crystalwas of the order of to 104 cm. Measurements based on the deter-mination of total surface area of powdered bone by gas-adsorption tech-niques have produced results supporting earlier estimates of size. Suchparticles are well beyond the range of the light-microscope and it is onlywith the aid of the electron-microscope that actual crystals have beenvisualised.3 Using a combination of microdiffraction of X-rays and electron-microscopy, Engstrom and Zetterstrom estimated that the crystals were200 A in width. Robinson and Watson 5 produced electron micrographs ofintact bone showing crystals 350400 A long, almost as wide, and 25-50 Athick.The crystals appeared to be oriented with their long axes in thedirection of the collagen fibres. It seems established, therefore, that bonemineral is crystalline and that the crystals are very small indeed.Composition of the Crystals.-The main mineral constituents of bone andteeth are calcium, phosphate, carbonate, hydroxyl ions, and water, withsmaller amounts of magnesium, sodium, potassium, and chloride.The main component of the solid phase is undoubtedly some form ofbasic calcium phosphate. If the major component were tricalcium ortho-phosphate Ca3(P0&, then the molar Ca : P ratio would be 1-5. Manyworkers have shown that the Ca : P ratios of samples of human bone are ingeneral higher than 1.5, the most usual values being around 1.66, but valuesabove and below this figure are quite commonly found. Dallemagne andFabry6 are of the opinion that bone salt may be considered as having afundamental unit which is tricalcium phosphate combined with excess ofcalcium.In view of the similarity of the X-ray-diffraction patterns of bone andnaturally occurring apatites, many workers have concluded that bonemineral is an apatite.This is, however, inconclusive evidence, since anyunignited precipitate of calcium phosphate gives the X-ray diffraction pat-tern of an apatite. Neuman and Neuman,' in an excellent review, discussthis problem fully and conclude that the apatite pattern of X-ray diffractionis given by almost any sample of calcium phosphate with a molar Ca : Pratio lying between 1.33 and 2.0.Thus, on this basis, any unignited calciumphosphate sample with a Ca : P within 20% of the theoretical value of 1-66could be classified as hydroxyapatite. It is not justifiable, therefore, to1 W. F. De Jong, Rec. Trau. chim., 1926, 45, 445; W. F. Bale, H. C. Hodge, andS. L. Warren, Amer. J . Roentgenol., 1934, 32, 369; J. Thewlis, Proc. Phys. SOC., 1939,51, 99.N. V. Wood, Science, 1947, 105, 531.A. Engstrom and R. Zetterstrom, Exptl. Cell. Res., 1951, 2, 268.R. A. Robinson and M. L. Watson, Anat. Bec., 1952, 114, 383.3 R. A. Robinson, J . Bone and Joiizt Surg., 1952, 34, A , 389.6 M. J. Dallemagne and C.Fabry, in Ciba Foundation Symposium on Bone Struc-7 W. F. Neuman and M. W. Neuman, Chem. Rev., 1953, 53, 1.ture and Metabolism, Churchill, London, 1956324 BIOLOGICAL CHEMISTRY.visualise the main component of bone salts as a compound of fixed com-position such as hydroxyapatite, 3Ca,(PO,),,Ca(OH),.Neuman and Neuman7 conclude that the fundamental bone salt is acompound of calcium, phosphate, hydroxyl ions, and water which exhibitsa Ca : P ratio of approximately 1.5 and diffracts X-rays to give a character-istic apatite pattern. It represents only one small region of an indefiniteseries, the transition from one end of which to the other is associated withisomorphic substitution of hydrogen ions and water for calcium. This viewis in accord with the concept of the dynamic state of the bone minerals.Fabry has introduced the term pseudoapatite to describe the fundamentalcompound of bone salt.The best estimate of the bone salt therefore seems to be that it consistsof a crystal lattice containing mainly calcium, phosphate, and hydroxyl ions,the outer components of which are in equilibrium with a surface hydrationshell containing calcium and other ions and also in equilibrium with theinterior ions of the crystal.The hydration shell in turn is considered to bein rapid equilibrium with the tissue fluids in the bone. This concept is notcompletely irreconcilable with the views of Dallemagne and his colleagues 8who have always opposed the idea that bone was essentially hydroxyapatite,nor does it jettison completely the idea of an " apatite-like " structure.Carbonate of Bone Mineral.-The mineral portion of bone contains about5% of carbon dioxide.One of the difficulties in the study of bone salts isthat the preparation of samples for investigation necessitates destruction ofthe organic phase; this gives rise to the possibility of producing changes inthe mineral fraction. Dallemagne and his co-workers have investigatedthe carbonate content of bone salts obtained from whole bone by the re-moval of the organic material with ethylene glycol. They consider thatcalcium carbonate has-an independent existence in bone salt, and that it isnot part of the crystal lattice. It has been shown that the thermal extrac-tion curve for carbon dioxide of bone salts is similar to that for decom-position of calcium carbonate.When bone salts are dissolved in dilute acids,carbon dioxide is released faster than phosphate is dissolved. In youngrats,lo experiments with l*CO, have shown that skeletal carbon dioxide is incomplete equilibrium with blood, which would be unlikely if the carbonatewere situated deep in the crystal lattice. These observations provided cir-cumstantial evidence for considering that the carbon dioxide is surface-bound but, as Neuman and Neuman7 point out, they do not necessarilyprove the independent existence of calcium carbonate. Dallemagne andFabry 6 proposed that basic tricalcium phosphate may be bound to carbondioxide via an additional calcium atom.Size of Bone Crystal.-Bone crystals are so small that they present a vastsurface area. Robinson,3 on the basis of average crystal dimensions, hascalculated that the specific surface in autoclaved bone varies from 84 to106 sq.m./g. Thus in an average man the total bone crystal surface would8 C. Fabry, Biochim. BioFhys. Acta, 1954, 14, 401.9 W. F. Neuman, T. Y . Toribara, and B. J. Mulryan, .J. Amer. Chem. SOG., 1953, 75,10 b. L. Buclianan and A. Nakao, Fed. PYOC., 1952, 11, 19.4239HARTLES: CHEMISTRY OF BONE AND TOOTH MINERALS. 326be of the order of 100 acres. The few litres of body fluid flowing over thesurfaces could therefore be in intimate contact with the solid phase, thusfacilitating the rapid exchange so often observed in physiological studies.Tooth Minerals.-The fundamental difference between the tooth of non-persistent growth and bone is that once the tooth is formed it does notundergo the biological remodelling which can take place in bone.Localfactors may cause small variations in the composition of teeth during theirformation and these variations will then persist throughout the life of thetoo th.llThe individual crystals in enamel and dentine are much larger than inbone.12 The total surface area presented by the crystals is therefore muchless, values of 1-8 and 2.4 sq. m./g. have been reported for enamel and dentine,respectively2 (cf. 84-106 sq. m./g. for bone). In general, however, thecomposition of the fundamental tooth salt is believed to be similar to thatof bone, but much work remains to be done on minor differences in com-position, particularly in relation to the problem of dental caries.There issome evidence to suggest that, in the cotton rat, teeth with a high carbonatecontent are more liable to decay than those with a lower carbonate content.11Influence of Diet on Composition of Bones and Teeth.-Severe changes inthe mineral content of the diet are reflected in the composition of bone ofany age, but are most noticeable in young growing bone. Dietary changescan only influence the tooth during its formative period. An importantseries of papers have been published by Sobel and his colleagues during thelast few years.11J2" They have shown that in the rat and the cotton ratthere is a relation between the composition of bone and tooth minerals andthe fluid from which they are deposited, and that the composition of the fluidis in turn related to that of the blood serum.Animals reared from weaningon a diet high in calcium and low in phosphorus bad a blood-calcium concen-tration 10% above normal, a blood carbon dioxide level about 10% belownormal, and a blood-phosphorus value only 40% of normal. On a diet lowin calcium and high in phosphorus the blood-calcium concentration was about60% of the normal, blood carbon dioxide about 4% above normal, andblood-phosphorus about 17% above normal. These changes in blood com-position were reflected in different ways in the different mineralised tissues.In the normal cotton rat the PO, : ZCO, ratio is highest in enamel, lower indentine, and lowest in bones.The Ca : PO, ratio of the tissues shows muchless variation, although in enamel it is still slightly higher than in dentine orbone.In the tibia and femur, and in the enamel and dentine of the incisor andmolar teeth, the PO, : XO, ratio is directly related to that of the serum.In contrast, the Ca : PO, ratio of enamel and dentine is hardly affected bychanges in blood Ca : PO, ratio, whereas the ratio in bone varies with that11 A. E. Sobel, Ann. N . Y . Acad. Sci., 1955, 60, 713.12 R. F. Sognnaes, D. B. Scott, M. J. Ussing, and R. W. G. Wyckoff, J . Dent. Bes.,1952, 31, 85.125 A. E. Sobel, M. Rockenmacher, and B. Kramer, J . Bid. Chem., 1945, 158, 475;1945, 159, 159; A. E. Sobel and A. Hanok, ibid., 1948, 176, 1103; A.E. Sobel, A.Hanok, H, Kirschner, and I. Fankuchen, ibid., 1949, 179, 205; A. E. Sobel and A.Hanok, J . Dent. Res., 1958, 37, 632326 HIOLOGICAL CHEMISTRY,in the serum. These resuIts refer to growing bones and growing teeth andindicate differences in composition related to differences in dietary intake ofcalcium and phosphorus. The major changes arising from alteration in dietare in the PO, : 2C0, ratios. The changes in Ca : PO, are smaller, thoughsignificant, in bone and not significant in teeth. These observations areinterpreted by Sobel as indicating the governing influence of " local factors,"acting a t the site of formation, upon the composition of the bone and tooth.It is also apparent that the "carbonate" fraction of the tooth is morereadily susceptible, to dietary influence than is the calcium or phosphateportion.This in turn suggests that the " carbonate " of the tooth is notincorporated in the crystal lattice but is situated at the surface or in thehydration cell.Moisture Content of Mineralised Tissues.-The water content of bonevaries with the source of bone and with its age. Human bone from thenewborn has been reported to contain 30% of water, compared with 20% inold age.13 Other workers have found values of 13-22y0 of water in humancortical bone and 32-52% in cancellus bone.14 The water content ofpowdered bone equilibrated in air for 48 hours and then heated for 24 hoursat 105" was found l5 to be 8%. Cortical bone from young dogs contains asmuch as 540h of water compared with 21% in older dogs.16Le Fevreand Manly l7 reported that enamel contained an average of 2.3% of water(range 1-5%) and that dentine had an average water content of 13.2%(range 10.8-15.7 %) .Deciduous enamel (2.8%) contains slightly morewater than permanent enamel (2.3%). Deciduous dentine, on the otherhand, contains less water (11.2y0) than permanent dentine (13.2%). In arecent study Burnett and Zenewitz l8 found that the maximum water con-tent of freshly extracted whole teeth and dentine was 9-32 and 10.Oyo,respectively. Rehydration a t 98" F and lOOyo humidity restored some, butnot all, of the moisture content.Earlier work has shown that minerals in fresh untreated bone, althoughcontaining considerable amounts of water, are not as highly hydrated as anequilibrated synthetic hydroxyapatite.This has led to the tentative con-clusion that not all the crystal surface of intact bone is available for hydr-ation, perhaps owing to its bonding with the organic phase.' From this itmay be inferred that with increasing age a larger proportion of adult compactbone may be in less intimate contact with water, either as a hydration shellor as circulating fluid. For example, it is now well established' that theuptake of 45Ca and 32P is greatest in young growing bone. This may dependa t least in part on the degree of hydration of the bone crystals.Association of Mineral Crystal with the Organic Phase.-Electron-micro-graphs 3~19 of intact sections of bone show an intimate contact between theIn general, the water content of teeth is less than that of bone.13 J .H. Vogt and A. Tonsager, A d a Med. Sand., 1949, 135, 231.14 I. S. Edelman, A. H. James, H. Baden, and F. D. Moore, J . Clin. Invest., 1954,33, ;522.J . E. Eastoe and B. Eastoe, Biochem. J., 1954, 57, 453.16 R. A. Robinson and S. R. Elliott, J . Bone and Joint Surg., 1957, 39, A , 167.17 M. L. Le Fevre and R. S. Manly, J . Amer. Dent. Ass., 1938, 25, 233.18 G. W. Burnett and J . Zenewitz, J. Dent. Res., 1958, 37, 581IIARTLES: CHEMISTRY OF BONE AND TOOTH MINERALS. 327suriaces of the crystals and the collagen fibres. The problem arises,what if any is the nature of the linkage between the mineral and organicphase?Robinson and Watson l9 review existing information and present con-vincing evidence for the association of mineral crystals with the bandregions of collagen fibrils of bone. In human infant bone, a small 100-120 Aperiod banding is observed.In mature bone the spacing alters and a pair ofbands appears at 640 The diameter of the fibril also increasesirom 150 A in very young infants to about SO0 A in middle-aged adults.The inorganic crystals are found in association with these bands. Thecrystals in young bone are very small, -100 A in length; in mature bonethey are larger, 200-300 A long, and span the doublet band of the fibril.In senile bone the diameter of the collagen fibril increases, to 1500 A, andthe crystals are large enough to spread over two or more doublet bands,thus obscuring the fibril period.The first appearance of inorganic crystalsis associated with the appearance of the collagen fibrils; there does there-fore seem to be an intimate link between the two structures. The differencein size of crystals from young and adult bone may be a further factor in thegreater isotope-exchange which occurs in newly formed bone.Role of Citrate.-Bone and dentine contain almost 1% of citrate, andenamel contains nearly 0.1yo.20 Little is known of the function of citratein a mineralised tissue. Dixon and Perkins 21 suggest that bone citrate isformed in bone cells by normal metabolic activity and is coprecipitated withminerals during calcification of the tissue. This is a reasonable suggestionsince a coprecipitate is formed from solutions of inorganic and citrate ionssimilar to those found in a serum ultra-filtrate.22 Armstrong and Singer 23believe that a t least some, if not the major portion, of bone citrate is ofpurely adventitious origin owing to the continuous presence of citrate inbody fluids.Bellin and Steenbock 24 found that administration of vitamin Dto previously depleted animals caused an increase in bone citrate but con-cluded that the amount of citrate in bones was related to the calcium nutri-tion of the animal rather than to the rachitic state per se. Nicolaysen andEeg-Larsen 25 consider that there may be a dual effect, that vitamin D doesinfluence the accumulation of citrate in bones but that over longer periods adefect arising from deficiency of the vitamin may be ameliorated when thediet is rich in calcium and phosphorus.When whole bone or dentine is treated with dilute hydrochloric acid theportion dissolving contains the minerals, about 5% of the nitrogenous com-pounds, and all the citrate (06--0.9%).Recent work has suggested theintervals.19 R. A. Robinson and M. L. Watson, Ann. N.Y. Acad. Sci., 1955, 60, 596.20 F. Dickens, Biochem. J., 1941, 52, 260; A. H. Free, J . Dent. Res., 1943, 22, 477;I. Zipkin and I<. A. Piez, ibid., 1950, 29, 498; M. V. Stack, Brit. Dent. J., 1951, 90, 173.21 T. F. Dixon and H. R. Perkins, Biochem. J., 1952, 52, 260.22 A. C. Kuyper, J . Biol. Chem., 1945, 159, 411.23 W. D. Armstrong and L. Singer, in Ciba Foundation Symposium on Bone Struc-24 S.A. Bellin and H. Steenbock, J . Biol. Chem., 1952, 194, 311; H. Steenbock and25 R. Nicolaysen and N. Eeg-Larsen, in Ciba Foundation Symposium on Boneture and Metabolism, Churchill, London, 1956.S. A. Bellin, ibid., 1953, 205, 988.Structure and Metabolism, Churchill, London, 1956328 BIOLOGICAL CHEMISTRY.association of citrate with a peptide.Z6 When dentine from human teeth isdemineralised the solution contains the minerals and the citrate-peptidecomplex. If minerals are reprecipitated by raising the pH, the citratecomplex is adsorbed or coprecipitated with the minerals. Lowering the pHto redissolve the minerals releases the citrate complex. Precipitation ofcalcium as sulphate at an acid reaction does not cause coprecipitation of thecitrate complex.Thus, over the physiological range of pH, citrate is firmlyassociated with the mineral phase. This linkage of citric acid with mineraland a peptide is of interest since it suggests that not all the citrate presentin a mineralised tissue is of adventitious origin. Analysis has shown thatthe peptide associated with the citrate is highly basic, containing a largeproportion of arginine and ammonia, with aspartic acid, valine, Ieucine, andisoleucine as major cornponent~.~~ There is, as yet, no evidence that thecomplex is associated with the collagenous constituents of dentine, unlessby ionic linkage.goStrontium and Mineralised Tissues,-The increased concentration ofbone-seeking alkaline-earth nuclides arising from nuclear fission provides arecent alteration in the general environment which requires careful study.Among these baleful products of man's ingenuity are B9Sr and 90Sr; thelatter is widely distributed in human bone and is potentially noxious owingto its comparatively long half-life of 28 yearsa8Stable strontium is a normal minor constituent of bone.ee Using animproved method of determination by radioactivation analysis, Sowdenand Stitch 30 found that samples of bone from normal persons of both sexesand different ages contained about 100 pg./g.of ashed tissue. There is noevidence that quantities of strontium of this order are harmful to bone.The possible hazard to health from ingestion of 90Sr would therefore be dueto its radiation activity and not to its chemical toxicity.QOStrontium is de-posited in the United Kingdom approximately in proportion to the rainfallin a given area. The greatest uptake of the nuclide is by vegetation of hillpastures where the soil is acidic and often deficient in calcium and phos-phorus.31 Hill sheep which graze on these pastures are therefore sensitiveindicators of the degree of contamination of a particular area.Biochemically, strontium behaves very like calcium,32 but there is adefinite discrimination against strontium in the presence of an adequatecalcium intake. Thus, calcium is preferentially absorbed from the gut,and strontium is more readily excreted via the urine than is calcium. Thereis therefore preferential utilisation of calcium in bone formation and inlactation.Two recent papers are of particular interest; Morgan and26 R. L. Hartles and A. G. Leaver, Arch. Oral Biol., 1960,1, in the press.27 A. G. Leaver, J, E. Eastoe, and R. L. Hartles, Arch. Oral Biol., 1960, 1, in the28 J. L. Kulp, W. R. Ecklemann, and A. R. Schubert, Science, 1957, 1245, 219;2s R. M. Hodges, N. S. MacDonald, R. Nusbaum, R. Steams, F. Ezmirlian, P. Spain,80 E. M. Sowden and S. R. Stitch, Biochem. J . , 1959, 67, 104.31 F. J. Byrant, A. C. Chamberlain, A. Morgan, and G. S. Spicer, J . Nuclear Energy,32 H. G. Jones and W. S. Mackie, Brit. J . Ntctr., 1959,18, 355.press.W. R. Ecklemann, J. L. Kulp, and A. R. Schubert, ibid., 1958, 12'9, 266.and C. McArthur, J . Bid. Chem., 1950, 186. 619.1967, 6, 22HARTLES: CHEMISTRY OF BONE ,4ND TOOTH MINERALS.329Wilkins 33 analysed the carcass of a yearling sheep reared on hill pasture inan area of high rainfall. The animal was killed in 1957, and analysis showedthat the average activity of the whole skeleton was 182 strontium units(1 S.U. = c per g. of calcium), the level in the teeth was lower (135 S.U.)and that in the pelvis was 203 S.U. Jones and Mackie 32 carried out experi-ments on Scottish Blackface wethers aged 12-15 months; they adminis-tered 89Sr and 45Ca simultaneously to their animals and found that theproportion of the dose absorbed and deposited in the skeleton was for 4Tafour times that for 89Sr. They found that the major discrimination against89Sr was in absorption and urinary excretion. There appeared to be littleor no discrimination in rate of excretion into the intestine or in transfer ofnuclides from serum to bone.An interesting experiment is reported by Holgate.= Rabbits were givena single injection of 90Sr c per 100 g.of body weight), and the uptakewas determined for teeth and femur, Rabbits have teeth which are con-tinuously growing, but they remain constant in length owing to attrition.The 90Sr content of the femur was maximal eight hours after injection andthen fell steadily to a minimum after 30 days, this level was then maintainedalmost constant until the last animals were killed at 180 days. This suggeststhat most of the 90Sr is rapidly adsorbed on the surface of the bone crystalsor stays in the hydration shell. Most of the 90Sr in the bone then returns tothe tissue fluids and blood as their content of %r falls.A smaller portionpenetrates to the interior of the crystal lattice, possibly by a recrystallisationprocess, and is only slowly removed as remodelling occurs, In teeth (con-tinuously growing) the picture was quite different. The gOSr in teeth in-creased steadily for 30 days after the injection and then fell to a smallervalue than in bone after 100 days. This can be explained by the liberationof 90Sr from bones into the blood stream in the days following injection, thusproviding a continuously available t hougb decreasing supply of nuclide forincorporation into teeth. Once formed, tooth mineral is much more resistantto change than is bone, and the gOSr burden remains until the tooth is wornaway by normal attrition.Thus the amount of 90Sr in the tooth willincrease until the first deposit reaches the biting surface and begins to wearaway. In an animal with teeth of non-persistent growth, such as themonkey or man, 90Sr taken up by the developing teeth after exposure to asingle high dosage is retained for a very long time. Thus a child exposed toa high level of 90Sr might expect to have radioactive teeth so long as thoseteeth remained in sit%, whereas the level in the bones would begin to decreaseshortly after exposure.Information concerning the 90Sr content of human bone is not easilyobtained; Holgate 34 quotes the annexed data as the latest available figures.1 month t o 1 year ..................... 0.70 S.U.1-1 S.U.1 year to 5 years ........................ 0.83 S.U. 1.2 S.U.5 years to 20 years 0.25 S.U. 0.45 S.U.Over 20 years ........................... 0.11 S.U. 0.1 S.U.Age March, 1956 July, 1956.................................... Stillborn 0.44 S.U. 0-55 S.U......................33 A. Morgan and J. E. Wilkins, Biochem. J., 1959, 71, 419.34 W. Holgate, Brit. Dent. J., 1959, 107, 131330 BIOLOGICAL CHEMISTRY.The largest rise is in children under 5 years of age. The Medical ResearchCouncil suggested that the burden of gOSr in human bone should not beallowed to rise above 100 S.U., and that if the level reached 10 S.U., then theproblem should receive immediate consideration.Status of Fluoride in Bones and Teeth.-Attention has been focusedrecently on problems associated with the skeletal deposition of fluoride sincemany communities are consuming water which naturally contains fluoride,and others are drinking water to which fluoride has been added to bring theconcentration up to one part per million (1 p.p.m.).The latter measurehas been adopted in certain areas in the United States and in three demon-stration areas in the United Kingdom, for it has been shown that the con-sumption of such a drinking water results in about a 50% reduction indental-caries experience. The topic of the relation of fluoride to dentalcaries has been extensively re~iewed.~5Fluoride is bone-seeking and all bones appear to contain some fluoride.Its concentration in the skeleton increases with advancing years and in somemeasure with dietary intake.36 Jackson and Weidmann 37 have recentlyexamined the fluoride content of human bone in relation to age and watersupply in three areas in the United Kingdom, where the water contained(0.5, 0.8, and 1.9 p.p.m.of fluoride respectively. They found that in allinstances the amount of fluoride in bone increased with age up to a maximumat about 55 years. At this point bones from the three areas contained 190,245, and 400 rng. per 100 g. of dry, fat-free material.In the case of teeth, fluoride is deposited systematically only during theirformation. It has been found that the fluoride content of enamel increaseswith the fluoride content of the water supply.37 Brudevold and his col-leagues 38 have suggested that fluoride deposition in enamel takes place inthree stages, during enamel formation, after mineralisation is complete butbefore eruption of teeth, and after eruption during the life span of the tooth.The last two methods of deposition are due to physicochemical changes inthe outermost layers of the formed enamel. The maximum concentrationof fluoride (3370 p.p.m.) was in the outer enamel from teeth formed in anarea where the drinking water contained 5.0 p.p.m. The correspondinginnermost enamel contained 570 p.p.m.Comparative figures from areaswhere the water contained 1.0 and 0.1 p.p.m. were 889 and 129 p.p.m., and499 and 42 p.p.m., respectively.The fixation of fluoride is believed to be by exchange with hydroxyl inthe crystal lattice or a t the crystal ~urface.3~ McCann,4O studying the up-take of fluoride by a synthetic hydroxyapatite over a wide range, concluded55 ’‘ The Fluoridation of Domestic Water Supplies in North America,” H.M.S.O.,London, 1953; H.H. spnes, Brit. Dent. J., 1954.96, 173; W. F. Stilwell, N. L. Edson,and P. V. E. Stainton, The Fluoridation of Public Water Supplies,’’ 1957, GovernmentPrinter, Wellington, N.Z. ; Wld. Hlth. Org., Techn. Rep. Ser., 1958, No. 146.s6 G. E. Glock, F. Lowater, and M. M. Murray, Biochem. J., 1941, 35, 1235.57 D. Jackson and S. M. Weidmann, J. Path. Bact., 1958, 76, 451.58 S. Isaac, F. Brudevold, F. A. Smith, and D. E. Gardner, J . Dent. Res., 1958, 37,SB W. F. Neuman, M. W. Neuman, E. R. Main, J. O’Leary, and F. A. Smith,40 H.G. McCann, J . Biol. Chem., 1953, 201, 247.318.J. Biol. Chem., 1950, 187, 655CLOVER : META~lROLISM OF P-CAROTENE AND RELATED PROVITA4MIKS A. 331that a t levels of a few parts per million fluorapatite was formed exclusively;with increasing concentration of fluoride a mixture of fluorapatite andcalcium fluoride was formed, until at concentrations above 0.2% calciumfluoride is the main product. The uptake of fluoride by the enamel surfacehas been studied by using lSF, and a method has been developed for theanalysis of small quantities of stable fluoride using an isotope-dilutiontechnique .*IFrom the dental point of view the important observations have beenmade that the solubility of surface enamel at pH 4 - 6 decreases as thefluoride content increases.42 Much more information is required concerningthe manner in which fluoride is deposited in a mineralised tissue.Conclusion.-The study of the chemistry of bone and teeth bears a directrelation to three major problems.First, in an ageing population where anincreasing number of individuals is surviving beyond three score years andten, consequences of bone fracture are serious; the chemistry of ageing andsenile bone must receive greater attention if the reasons for the slow healingof fractures of aged bone are to be understood. Secondly, the incidence ofdental caries in young children is distressingly high ; an increased knowledgeof the intimate chemistry of the tooth may provide information concerningthe decay process. Thirdly, the recently increased concentration of bone-seeking nuclides in the general environment requires that the pattern of theirskeletal deposition be studied.R.L. H.3. METABOLISM OF JS-CAROTENE AND RELATED PROVITAMINS AALTHOUGH 30 years have passed since Moorel demonstrated that p-caro-tene (1) was transformed into vitamin A (13) in the animal body, the mech-anism of this process represented by route A of Chart 1 is still unknown.The structural relation of the two compounds suggested to Karrer and hisco-workers 2 that p-carotene might undergo hydrolysis of the central doublebond to form two molecules of vitamin A, but attempts to elucidate thedetails of the biological system have been fruitless. Consequently, parallelwork on the structural features required for provitamin A and vitamin Aactivity has received more attention.This has been reviewed by Heilbronet aZ.,3 Ze~hmeister,~ B a ~ t e r , ~ Goodwin,6 and, more recently, by Isler andZeller.’The chemistry and biochemistry of the various carotenoids and related4 1 J. H. Fremlin, J. L. Hardwick, and J. Suthers, Nature, 1957, 180, 1,179; J. L.p 2 S. B. Finn and C. de Marco, J . Dent. Res., 1956, 35, 185; S. Isaac, F. Brudevold,Hardwick, J. H. Fremlin, and J. Mathieson, Brit. Dent. J., 1958,104, 47.F. A. Smith, and D. E. Gardner, ibid., 1958, 3’4, 254.T. Moore, Biochem. J., 1930, 24, 692.2 P. Karrer, R. Morf, and K. Schopp, Helv. Chim. Ada, 1931, 14, 1036, 1431.3 I. M. Heilbron, W. E. Jones, and A. L. Bacharach, Vitamins and Hormones, 1944,4 L.Zechmeister, Vitamins and Hormones, 1949, 7, 57.J. G. Baxter, in ‘ I Progress in the Chemistry of Organic Natural Products,” 1952,T. W. Goodwin, “ The Comparative Biochemistry of the Carotenoids,” Chapman &2, 156.Vol. IX, p. 41 (ed. L. Zechmeister), Springer-Verlag, Vienna.Hall Ltd., London, 1952.7 0. Isler and P. Zeller, Vitamins and Hormones, 1957, 15, 31332 BIOLOGICAL CHEMISTRY.compounds have also been described in considerable detail by Karrer andJucker * and G o o d ~ i n , ~ , ~ respectively. Again, work on the chemicalsynthesis of vitamin A and various carotenoids and larger homologues hasalso been reviewed r e ~ e n t l y , ~ ~ * J ~ J ~ but detailed knowledge regarding theirmetabolism is still lacking. By newly developed methods of chemicalsynthesis, a variety of compounds, intermediate in size between P-caroteneand vitamin A, can be prepared which should assist the biochemist indetermining the nature of enzymic attack on molecules such as carotenoidswhich possess long, conjugated double-bond systems. The present Reportdiscusses work carried out during the last few years towards this end withparticular reference to the provitamin A-vitamin A transformation.( 1 ) 4 f\1- [Unknown i n t e r m e d i a t e dCHART 1.Suggested routes for eonvevsion of @-carotene into vitamin A .* P. Karrer and E. Jucker, “ Carotenoids,” transl. E. Braude, Elsevier, Amsterdam,@ T. W. Goodwin, Ann. Rev. Biochfm., 1955, 24, 497.10 H. H. Inhoffen and H. Siemer, Progress in the Chemistry of Organic Naturall1 0.Isler, H. Lindlar, M. Montavon, R. Riiegg, G. Saucy, and P. Zeller, Chem. SOC.1950.Products,” 1952, Vol. IX, p. 1 (ed. L. Zechmeister), Springer-Verlag, Vienna.Special Publ. No. 4, 1956, p. 47GLOVER: METABOLISM OF @-CAROTENE AND RELATED PROVITAMINS A. 333Provitamin A-Vitamin A Conversion.-Structurally, it would appear that,if P-carotene were oxidised a t the central double bond, two molecules ofretinene (vitamin A aldehyde) (11) might be formed which could be imme-diately reduced to vitamin A.12 On the other hand, if an excentric bondis attacked, only one molecule of vitamin A would result from the furtherdegradation of the larger fragment (see Chart 1). The evidence for andagainst these two views has been outlined p r e v i o u ~ l y .~ ~ ~ ~ ~ ~ ~ A major diffi-culty of the problem is that the conversion of p-carotene into vitamin Atakes place only on a small scale in experimental animals and relativelyslowly. Further, it has not yet been possible to prepare an enzyme systemwhich will carry out the reaction in vitro, so the possibility of obtainingsufficient intermediates to allow their proper characterisation seems remote.An alternative approach is to synthesise substances closely related tobiological intermediates indicated by some hypothetical scheme such asterminal fission followed by p-oxidation (route B in Chart 1). Here eitherthe fhapocarotenals (2, 5, 7, and 9) or the related series of p-apocarotenoicacids (4, 6, 8, and 10, respectively) would be possible intermediates.Therewere several reasons for considering this possible, e.g. : (a) Most biologicalassays l5 suggest that only one molecule of vitamin A is formed per moleculeof p-carotene, although perhaps two may be obtained in the presence ofoptimal amounts of to~opherol.l6~~~ No intermediate values have beenobtained. (b) Two yellow pigments were isolated18 from the lipids of thehorse intestine and tentatively identified 19 as p-apo-10’- (5) and p-apo-12’-carotenal (7) ; these must have resulted from terminal oxidation ofp-carotene. They had not hitherto been detected ih plant extracts, so itwas assumed they were formed in the animal. (c) Chemical oxidation of@-carotene begins at one end of the conjugated double-bond system {seebelow) , and often there is an overall parallelism between chemical and bio-logical oxidations.(a) 16,16’-Bishomo-~-carotene (14) is biologicallyactive 20 (20% as active as all-trans-@-carotene) although it does not possessa central double bond. (e) Substances such as a-“ vitamin ” A 2 l or 3-hydr-oxy-“ vitamin ” A z2 which would arise in addition to vitamin A fromla R. F. Hunter, Nature, 1946, 158, 257; J. Glover, T. W. Goodwin, and R. A.l a J. S. Lowe and R. A. Morton, Vitamins and Hormones, 1956, 14, 97.14 T. Moore, “ Vitamin A,” Elsevier, Amsterdam, 1957.15 F. M. M. Hume, Brit. J . Nutrit., 1951, 5, 104.l6 C. J. Koehn, Arch. Biochem. Biophys., 1948, 17, 337.1’ M. J. Burns, S. M. Hauge, and F. W. Quackenbush, Arch. Biochem.Biophys.,18 G. N. Festenstein, Ph.D. thesis, Liverpool, 1951.Is E. R. Redfearn, Ph.D. thesis, Liverpool, 1954.Zo H. J. Deuel, jun,, H. H. Inhoffen, J. Ganguly, L. Wallcave, and L. Zechmeister,21 S. R. Ames, W. J. Swanson, and P. L. Harris, J. Amev. Chew SOL, 1955, 77, 4136.22 R. H. Painter, Ph.D. thesis, Liverpool, 1955.Morton, Biochem. J., 1948, 43, 109.1951, 30, 341.Arch. Biochem. Biophys., 1952, 40, 352334 BIOLOGICAL CHEMISTRY.central fission oi dietary or-carotene or cryptoxanthin, respectively, and areknown to be stored in the liver, have never been detected in rats.p-Apo-B’-carotenal (2) had been reported 23 to be vitamin-A-active, sothe remaining members of the series were prepared by oxidising p-carotenewith hydrogen peroxide in the presence of osmium tetroxide.24 Morerecently they have been elegantly synthesised by Ruegg and his col-league~.~~.26In studying the metabolism of p-apo-8’-, -lo’-, and -12’-carotenal in therat, it was observed19 that small amounts of some were oxidised to thecorresponding carboxylic acids ; consequently the higher vinylogues ofvitamin A acid were also prepared.Preparation of p-Apocarotenals-Oxidation of p-carotene. When p-caro-tene is oxidised with chromium trioxide, the double bonds of the @-iononerings are preferentially attacked and the end products are semi-p-carotenoneand p-~arotenone.~’ Alkaline permanganate, on the other hand, appears toattack the terminal double bonds of the central chain yielding p-apocaro-tenals (2, 5, 7) having one @-ionone ring intact, but it does not appear toform retinene 28 (1 1).With hydrogen peroxide alone, however, retinene isformed in small yield 29 (ca. 1%), though with osmium tetroxide as cata-lyst 24930 a moderate yield (cn. 30%) i: obtained. When a solution of hydro-gen peroxide in t-butyl alcohol is used,31 the progress of the reaction couldbe followed with time and the p-apo-8’-, -lo’-, and -12’-carotenal as well asretinene were isolated in chromatographically pure form.lg Grob andButler 32 also reported finding these aldehydes. More recently a smallamount of the P-apo-l4‘-carotenal (9) has also been found% in the reactionmixture, Dialdehydes, having 3,4, 5, and 6 ethylenic bonds in conjugation,and corresponding to segments of the central chain of p-carotene, are alsopre~ent.1~ The pattern of the rates of production of the various aldehydesby osmium tetroxide-hydrogen peroxide pointed to a progressive removalof the ends of the conjugated system rather than random reaction along it.19The reagent attacks the penultimate double bond in the conjugated system;semi-p-carotenone or P-carotenone have never been detected among theproducts.Further, the yields of pure retinene or higher aldehydes fromp-carotene were found never to exceed 10-12y0.In one procedure (Chart 2) for the synthesis of p-apo-12’-carotenal (C25) (7), Ruegg and colleagues 25 used as starting point theDirect synthesis.2s H. von Euler, P. Karrer, and U. Solmssen, Helv. Chim Acta, 1938, 21, 211.24 N.L. Wendler, C. Rosenblum, and N. Tishler, J . Amer. Chenz. SOC., 1950, 72,25 R. Ruegg, H. Lindlar, M. Montavon, G. Saucy, S. F. Schaeren, U. Schwieter, and26 R. Riiegg, M. Montavon, G. Ryser, G. Saucy, U. Schwieter, and 0. Isler, Helv.27 R. Kuhn and H. Brockmann, Annulen. 1935, 516, 95.28 P. Karrer and U. Solmssen, Helv. Chim. Acta, 1937, 20, 682; P. Karrer, U.29 R. F. Hunter and N. E. Williams, J., 1945, 554.30 G. C. L. Goss and W. D. Macfarlane, Science, 1947, 106, 375.31 N. A. Milas and S. Sussman, J . Amer. Chem. SOC., 1936, 58, 1302.32 E. C. Grob and R. Butler, Helv. Chinz. Acta, 1954, 37, 1908.33 U. Liithi, M. J. Fishwick, and J. Glover, unpublished work (1957).234.0. Isler, Helv. Chim. Acta, 1959, 42, 847.Chim. Acta, 1959, 42, 854.Solmssen, and W.Gugelmann, ibid., p. 1020GLOVER : hIET.4I30LISM OF @-CAROTENE .4ND RELATED PKOVI‘TAMINS A. 336C,, p-aldehyde (15) which is an intermediate in the industrial synthesis ofp-carotene. This was condensed with lithium acetylide in liquid ammoniato form the C,, acetylenic alcohol (16). The latter without purification was(19)CHART 2. Synthesis of ~-apo-l2’-caroteizul.coupled with methylmalondialdehyde enol-benzoate (17) in a Grigiiardreaction, giving the ester (18), which with acetic acid in propan-2-01 undernitrogen yielded 15,15‘-didehydro-~-apo-12’-carotenal (C,,) (19). Partialhydrogenation with a Lindlar catalyst 35 and isomerisation of the 15,15’-HO - Y C H ( O E t ) ,H C ~ C L ~ + OHCyCH*OEt + 9 C ; ; P E t(20) (22)(15)CHO - CH (OEt)j R+ CH (OEt),4- H,C:CH.OEt (+ZnC12)(25) 6R Y C H ( O E t )(2 4)CHART 3.Synthesis of 15,15’-didehydro-/3-apocarotenals.cis-p-apocarotenal intermediate produced a good yield of the all-trauts-p-apo-l2’-carotenal (7).Sodium or lithiumacetylide was first condensed with the methylmalondialdehyde enol-ether(20) to form the ether (21). This with ethyl orthoformate produced the34 0. Isler, H. Lindlar, 31. Montavon, R. Ruegg, and P. Zeller, Helv. Chinz. A r f a ,1956, 39, 249.35 H. Lindlar, Hrlu. CJCiwE. ilcfn, 1952, 35, 446.An improved procedure was followed later (Chart 3)336 BIOLOGICAL CHEMISTRY.acetal (22) which is a useful new building unit for branched polyenes. Con-densation of C,, p-aldehyde and this compound by means of lithium amidein liquid ammonia afforded a C,, hydroxy-acetal (23) which was readilyconverted by acid into the free aldehyde (19).The higher vinylogues were prepared 26 by using this aldehyde as startingmaterial. The diethyl acetal (24) was condensed with ethyl vinyl ether inthe presence of zinc chloride, to form the dehydro-p-C,, ether acetal (25),which on acid hydrolysis gives the free aldehyde in good yield.Repetitionof these steps using alternately ethyl propenyl ether and ethyl vinyl etherfor the condensation with the appropriate 15,15’-didehydro-p-apocarotenalacetals enabled Riiegg and colleagues 26 to synthesise the higher membersof the series up to C4,,. Reduction of the acetylenic bonds in each of thesevarious 15,15’-didehydro-~-apocarotenals with a Lindlar catalyst, followedby isomerisation of the cis-derivatives by heat, gave the all-trans-p-apo-carotenals which are listed with their ultraviolet absorption maxima inTable 1.Further reduction of the aldehydes with lithium aluminiumhydride affords the corresponding series of p-apocarotenols.TABLE 1. Ultraviolet absorption maxima (m p) of p-apocarotenoids inC0,Melight petrolewt (for R see Chart 1).Compound X=CH,*OH CHO CO,H326 367343, 355 385377, 393 414376, 393 410403, 424 435426, 453 457443, 471 473a Ref. 36.350376 373 li325-400327 @430 426 28448 445, 471 26458, 495 464, 491 28Synthesis of p-Apocarotenoic Esters.-Two of this series, ethyl p-apo-14’-carotenoate (28) and p-apo-l2’-carotenoate (31) containing 22 and 25carbon atoms respectively, were synthesised from retinene 36 (see Chart 4).In a modification of the Reformatsky reaction,37retinene was condensed with ethyl bromoacetate in refluxing pyridine-benzene (5% vlv) in the presence of zinc dust to form the C,, 15-hydroxy-ester (27) in 80--85~0 yield.Dehydration with freshly prepared aluminiumEster of the C,, acid.38 S. Fazakerley, Ph.D. thesis, Liverpool, 1957.37 R. L. Shriner, “ Organic Reactions,” John Wilcy & Soils Inc., New York, 1042,VOl. IGLOVER: METABOLISM OF p-CAROTENE AND RELATED PROVITAMINS A. 337phosphate then gave a variable yield (30-60y0) of ethyl P-apo-l4’-caro-tenoate.Reduction of the ester (28) with lithium aluminiumhydride formed the C,, alcohol (29) which was readily converted into thealdehyde P-apo-l4’-carotenal (30) by manganese dioxide.38 A similar con-densation of the latter with ethyl bromopropionate followed by dehydration,again with aluminium phosphate, enabled ethyl P-apo-12’-carotenoate (31)to be obtained.Condensation of retinene with ethyl y-bromotiglate wasused in the preparation of the 15-hydroxy-ester (32), but dehydration of thiscompound always yielded the retro-acid, a difficulty often experienced inchain-lengthening of polyene~.~~Ester of the C,, acid.CHART 4. C,, and C,, /%apocaroterioids.The isomeric C,, acid (33) having a methyl group in the F-position to thecarboxyl group was synthesised by Redfearn 40 using a procedure similar tothat of Robeson et aL4I for vitamin A.It involved the condensation ofretinene with methyl P-methylglutaconate to form a diester which on hydro-lysis and decarboxylation gives the C,, mono-acid (33).C,, to Cm acids. The higher vinylogues containing 27-40 carbons (e.g.,torularhodin) have been recently synthesised by Isler and colleagues 42by the Wittig reaction.& The chain-lengthening was carried out as follows :methyl bromoacetate and triphenylphosphine gave the phosphoniumbromide which with sodium methoxide or aqueous sodium hydroxide yieldedthe phosphorane (34). This compound reacted smoothly with 15,15’-di-dehydro-P-apo-lZ’-carotenal (C,J (19) ; the resulting ester (35) was partiallyhydrogenated and isomerised as described above, giving methyl all-trans-P-apo-10’-carotenoate (36) (C27).Saponification liberates the free acid.The higher members, e.g., p-apo-8’-carotenoic acid (39), were prepared byusing the same sequence of reactions with the various 15,15’-didehydro-P-apocarotenals and the appropriate phosphorane (34 or 37) (Chart 5).38 S. Ball, T. W. Goodwin, and R. A. Morton, Biochem. J . , 1948, 42, 516.39 0. Isler, W. Huber, A. Ronco, and M. Kofler, Helv. Chim. Acta, 1947, 30, 911;H. 0. Huisman, A. Smit, S. Vromen, and L. G. M. Fischer, Rec. Trav. chim., 1952, 71,899.4 O E. R. Redfearn, 1957, personal communication.4 1 C. D. Robeson, J. D. Cawley, L. Weisler, M. H. Stern, C. C. Eddinger, and A. J.42 0. Isler, W. Guex, R. Hiiegg, G. Ryser, G. Saucy, U. Schwieter, M. Walter, and43 G.Wittig and U. Schollkopf, Chem. Bcr., 1054, 87, 1964; G. Wittig and W. Haag,Chechak, J . Amer. Chem. Soc., 1955, 77, 4111.A. Winterstein, Helv. Chim. Acta, 1959, 42, 864.ibid., 1955, 88, 1664338 BIOLOGICAL CHEMISTRY.The absorption maxima of the series of aldehydes, acids, esters, andalcohols are given in Table 1.Provitamin-A Activity of p-Apocarotenoids.-The availability of the abovecompounds has enabled the terminal oxidation hypothesis to be examined.Preliminary results with the @-apocarotenals suggested to Glover and Red-fearn 44 that if oxidative fission of the 7',8'-double bond of p-carotene occurredTie., fission at (b) in formula (l)], the larger p-apo-8'-carotenal fragment mightbe degraded by @-oxidation. The 9'- and 13'-methyl groups lie in or-positionsto the potential carboxyl groups; they would slow the process but not stopit.It would, however, be stopped at the central carbon atoms because ofmethyl substitution in the @-position, at C(13). Again, if the normal p-oxid-ation were involved, one might also expect the activated forms of the p-apo-carotenoic acids to be intermediates.l e(3 9 )CHART 5 . Synthesis of ,!l-apocarotenoic acids.The metabolism of these synthetic compounds in rats has been ex-amined 44245 in two ways. First, the minimum daily requirement to promotesteady growth of vitamin-A-deficient rats at a rate at least equal to thatproduced by 1 pg. of vitamin A was determined under standard dietaryconditions. Secondly, lipids mainly from livers and intestines of groupsof vitamin-A-deficient rats dosed with the various P-apocarotenoids wereexamined for metabolites, in particular for vitamin A.I t is necessary todo this because growth-activity could mean merely conversion into vitamin Aacid and not vitamin A alcohol, which is the main product of @-carotenemetabolism. Not all the p-apocarotenoids have yet been tested in this way,but the results for some of the lower members of the series are set out inTable 2, together with those for vitamin A and its aldehyde and acid, as \Yellas for p-carotene.Recently, the biological activity of the @-apocarotenals containing25-32 carbons has been determined by Maurisch et al., using the U.S.P.rat curative growth assay.46 The results, which are included in Table 2, givea more precise measure of activities in relation to that of P-carotene.It is clear that all the compounds examined are vitamin-A-active, but44 J.Glover and E. R. Redfearn, Biochem. J., 1954, 58, xv.a5 S. Fazakerley and J. Glover, Biochem. J., 1967, 65, 3 8 ~ .46 Personal communication from Dr. M. Montavon (Messrs. Hoffmann-La Roche,Switzerland), regarding results reported by W. Maurisch, E. de Ritter, J. Vreeland,and R. Krukar at an Amer. Chem. SOC. meeting, September, 1959GLOVER: METABOLISM OF p-CAROTENE AND RELATED PROVITAMINS A. 339to varying degrees, confirming the earlier work of von Euler et aZ.23 Allthe compounds, except vitamin A acid, were converted in vivo into vitamin A,which was characterised by its chromatographic behaviour on alumina,ultraviolet absorption, and antimony trichloride colour r e a c t i ~ n .~ $ * ~TABLE 2. Provitamin A activity of some p-apocarotenoids in the rat.Biological activityfi-Apocarotenoid *c2() -R/CH2'OHR/CHO(vitamin A)/ CO2H Rp-CaroteneRef.a12b44--1944-494919---Dailyrequire-ment11123050-1005-1021< 2(tLg.)-5-10--5-0 Y* For R see Chart 1.'Vitamin Aformation all-trans-Activity (%) 46 v zDose15(mg. /rat)1.7-8.010-2.52.50.9{ ;:;2-51-101-101.60.6-1 4A7 fi-carotene -VitaminA (yo) Mean40-7060-700434 126-0.5 (24 hr.)7.0 (60 hr.)2130.2Trace3 d-10-15f P = 0.05.5.87240Range t86-18248-7065-8029-58a E.Le B. Gray, K. C. D. Hickman, and E. F. Brown, J. Nutrition, 1940, 19, 39.# I . M. Sharman, Brit. J. Nutrition, 1949, 3, viii. Ref. 23. Ref. 19.Whether the minimum daily requirements or the activities relative tothat of p-carotene are compared, the C,, compounds are superior to the othermembers so far examined and as good as or better than 6-carotene in sup-porting growth; so they could be related to intermediates in the metabolismof p-carotene or of the higher vinylogues. However, the low biologicalactivity of the C,, group, varying from one-fifth to one-fiftieth of the C,group, means that they are unlikely to be intermediates in the metabolismof the latter. The absence of any C,, compound in the lipids of animalsdosed with the various C,, compounds tends to confirm this.The lowe340 BIOLOGICAL CHEMISTRY.biological activities for the higher p-apocarotenals (C2, to C32) implies thatthey also are not intermediates in the main pathway for the formation ofvitamin A from p-carotene: however, they could be intermediates in aminor route.With regard to the ability of the various p-apocarotenoids to producevitamin A in vivo after single doses, only the ester of the C,, acid was superiorto p-carotene. When, therefore, the growth tests and the speed of conver-sion into vitamin A are taken into account, only the C25 15-hydroxy-acidreally behaves as if it were closely related to an intermediate on the mainroute of conversion of p-carotene into vitamin A.All these aldehydes appear to bewell absorbed by the rat.A portion of each is immediately reduced to thecarotenol in the intestine and another portion is oxidised to the acid; someof the latter then becomes esterified. In the lipids from rats dosed withp-apo-10‘-carotenal, a little p-apo-l2’-carotenol was found. Its formationmust be somewhat analogous to that of vitamin A alcohol from the esterof the C,, acid (see below). However, no compound with a carbon skeletonintermediate in size between either p-apo-8’-carotenal or p-apo- 12’-carotenaland vitamin A was detected in the lipids from animals dosed with thosealdehydes.Both these esters are well ab-sorbed and are stored in the fat depots as well as in the liver, whereas thefree acids tend to be metabolised quickly.This difference was also notedby Redfearn ** in studies with the C25 p-apo-l2’-carotenoic acid isomer (33)and vitamin A acid, and their esters: these three cannot be reduced to thecorresponding P-apocarotenols in vivo. Some of the C2, acid is convertedinto vitamin A in the intestine during absorption and the process iscompleted in the liver. If p-oxidation were involved in this step, thenvitamin A acid would be the main product and not vitamin A alcohol; soprobably a different enzyme system is utilised. Similuly, the C, acid oralcohol has not been detected among the metabolites of the ester of the C,acid.Chromatography of the recovered @-apo-IY-carotenoic ester fractionfrom the tissue lipids resolved it into two zones, both containing esters.Onecorresponded to the compound administered, having Amax. 396 mp, but thesecond had A,, 400 mp. This bathochromic shift is characteristic of thechange of a cis-compound into the more stable all-trans-f0rm,~*47 involvingthe double bond adjacent to the carbonyl group. About 40% of the totalester fraction was in the trans-form in the wall of the small intestine, whereasin the liver this proportion had risen to 93-94y0, indicating a progressivechange to the all-trans-f~rm.~~ This change probably accompanies hydro-lysis and re-esterifitation of the acid with either glycerol or a higher alcohol..The C,, acid isomer behaved similarly. Whether this change is necessarybefore conversion into vitamin A or merely coincidental is not known.Themetabolism of the higher P-apo-10’- and -8’-carotenoic acids or esters hasnot yet been fully investigated.47 H. 0. Huisman, A. Smith, P. H. van Leeuwen, and J, H. van Rij, Rec. Truu. chirra.,1956, 76, 977.p-A$o-8’-, -lo’-, and -12’-carotenaZs.p-Apo-14’- and -12‘-carotenoic estersGLOVER: METABOLISM OF @-CAROTENE AND RELATED PROVITAMINS A. 341The C,, isomer was studied 48 to determine the effect of methyl-substitu-tion in the @-position to the terminal carboxyl group on conversion of thecompound into vitamin A. Metabolism of this compound by the @-oxid-ation system is blocked, yet vitamin A alcohol is formed in small amounts.The yield was only slightly less than that from the normal @-apo-l2’-carot-enoic ester; so another enzyme system must be operative.Both the C,, (27) and the C,, acid (32) with a 15-hydroxyl group were converted into vitamin A, the latter in a yield (21% ofvitamin A in 19 hr.after a single dose) which is greater than can be obtainedfrom a single dose of P-carotene.As the C25 group of compounds proved the most active of the seriestested, they were examined further for participation in the p-carotene-to-vitamin A transformation. p-Apo-l2’-carotenal was selected first as repre-sentative of the group because it is readily converted into the acid andalcohol and would probably have the best chance to enter the enzymesystem responsible for metabolising @-carotene. The aim of the work wasto use the aldehyde as a trapping agent for any similar radioactive aldehydederived from the metabolism of specifically labelled [15,15’-14C]-@-carotenein the rat.Metabolism of [15,15’-14C]-@-Carotene.-Specifically labelled p-car0 tenewas synthesised 49 and a small sample * was mixed with an excess of @-apo-12‘-carotenal and administered to vitamin-A-deficient rats.The animalswere killed 5 hr. after dosing, when absorption across the intestine would beoptimal, and the lipids from the intestine and liver were examined.50 Un-changed @-apo-l2’-carotenal and @-apo-l2’-carotenoic acid were isolated andpurified chromatographically. The fraction containing the aldehyde wastreated with hydroxylamine and the oxime crystallised. The acid fractionwas too small for crystallisation even as a derivative, but it was purified byquantitative conversion into the ester followed by either rechromatographyor reduction with lithium aluminium hydride, the resulting alcohol beingisolated chromatographically. The specific activities of the various frac-tions were measured and compared with those of the original [15,15’-l4C]-P-carotene and the [1-14C] vitamin A isolated from the liver.The re-isolatedp-apo-l2’-carotenal and its metabolite, the corresponding acid, were labelled,indicating that a little @-carotene had been degraded to the aldehyde. Asthe aldehyde and acid had approximately the same specific activity, theformer must have become labelled first since the action of aldehyde-oxidaseis irreversible. Again, the agreement in their specific activities suggeststhat little or no [15,15’-14C]-@-apo-12’-carotenoic acid could have beenderived from a higher homologue by @-oxidation. The aldehyde re-isolatedfrom the lumen was radioactive, so presumably the [15,15’-14C]-P-carotenewas first attacked there.Comparison of the specific activities of the twocompounds shows that the amount formed was quite small (<3% of the48 E. R. Redfearn, Biochem. J., 1957, 66, 3 9 ~ .49 H. H. Inhoffen, U. Schwieter, C. 0. Chichester, and G. Mackinney, J. Amer.60 J. Glover and P. P. Shah (1958), unpublished work.* The Reviewer is grateful to Professor G- Mackinney for the gift of this material.15-Hydroxy-esters.Chew. SOC., 1955, 77, 1053342 BIOLOGICAL CHEMISTRY.dose). The specific activity of the [l-14C]-vitamin A was higher than thatexpected from the specific activities of the [15,15’-14C]-labelled p-apo-carotenoids and so must have been derived mainly by another route.Thisresult confirms earlier observations that, while a small amount of P-carotenemay be metabolised via carotenals, this is certainly not the main route.Metabolism of [U-14C]-p-Carotene and [U-14C]-Retinene.-An attempt 5lwas made to answer the problem of central fission versus terminal oxidationby comparing the metabolism of [U-l*C]-p-carotene and -retinene duringtheir absorption across the intestine of the rat. It was considered thatoxidative attack on a terminal bond in P-carotene, followed by degradationof the larger fragment to retinene or vitamin A, would lead to small fragmentswhich on entering the various metabolic pools would rapidly release 14C0,into the respired air.Central fission, on the other hand, would produce[U-14C]-retinene or [U-14C]-vitamin A directly, which being quickly ab-sorbed would perhaps not be appreciably metabolised until they reached theliver later. In this event, the pattern of release of 14C0, would be moregradual.[U-f4C]-p-carotene has been prepared 52953 in two laboratories and itsmetabolism studied in the rat. It was found to be degraded extensively tosmall fragments; 14C02 appeared quickly in the respired air, the amountbeing maximal 5 hr. after dosing (when absorption across the intestine isoptimal) and declining. This was consistent with the above hypothesis.However, when [U-14C]-retinene was used,64 the pattern of release of14C02 was almost identical, indicating that this molecule can also be rapidlydegraded in transit across the intestine.Conclusions.-(a) The general conclusion is that terminal oxidation ofp-carotene, followed by progressive removal of the side chain of the largerfragment until retinene or vitamin A is left, is not the main metabolicroute to vitamin A.(b) When such an oxidation does occur, however, thelarger fragment can give vitamin A, but not necessarily by p-oxidation.(c) The very strikingly greater biological activity of the C,5 p-apocarotenoidsthan of their homologues indicates that the former best provide the type ofsubstrate which the enzyme system requires for fission. The relatively lowyield of vitamin A obtained after single doses of P-apo-l2’-carotenal, com-pared with the biological activity at low dose level, is difficult to understand.It may be that when it is administered in small doses the yield of vitamin Ais better; or that vitamin A acid is produced to an appreciable extent. Thisacid would be difficult to trace chemically in vim, but is biologically active.J.G.Note added in proof. An exhaustive study by Worker 5 5 with whole organs andtissue preparations to find a suitable system for studying the conversion of /?-caroteneinto vitamin A has had little success.51 M. J. Fishwick and J. Glover (1958), unpublished work.52 J..Glover and E. R. Redfearn, Biochem. J., 1953, 54, viii; J. Glover and P. P.53 J. S. Willrner and D.H. Laughland, Can. J . Biochern. Physiol., 1957, 35, 819.54 M. J . Fishwick, Ph.D. thesis, Liverpool, 1958.65 N. A. Worker, Brit. J . Nutrition, 1959, 18, 400.Shah, abzd., 1957, 67, 1 5 ~ PERRY THE CHEMISTRY OF MUSCLE CONTRACTION AND RELAXATION. 3434. THE CHEMISTRY OF MUSCLE CONTRACTION AND RELAXATIONIN view of the long interval since any aspect of muscle biochemistry has beenreviewed in these Reports1 the present discussion will refer to recent ad-vances in the field set against the background of relevant earlier work.Contraction.-Structzcre of contractile system. Like many biologicalprocesses the chemical events associated with contraction take place in aheterogeneous phase. Presumably, metabolites and cofactors, both ofwhich are soluble and of relatively low molecular weight, interact with theinsoluble contractile system within the cell, causing it to shorten and tobring about contraction of the tissue as a whole.To understand thisZA1 P(4) (b) (aFIG. 1. Structure of the myofibril.(a) .4ppearance of myofibril of the striated muscle cell as seen with the ordinary light-(b) Diagrammatic representatio.pz of longitudinal section through myofibril i n the direction(c) Diagrammatic representation of a cross-section of the A band showing the hexagonalNote that the appearance of a longitudinalmicroscope. Diameter of myofibril, 1-2p.A-A (Fig. 1 6 ) .array of thick (A) and thin (I) filaments.section depends on the angle of sectioning.process detailed knowledge of the chemistry of the myofibril, the contractileunit, is not in itself enough; an understanding of the ultrastructure down tothe molecular level is also required.When examined in the light-microscopethe myofibril from striated muscle is characterised by dark and light bands,the A (anisotropic) and I (isotropic) bands respectively, which alternatealong its length (Fig. 1). Electron-microscopy has proved an invaluabletool for studies of ultrastructure and in the case of the striated myofibriltwo main filament types can be identified. The larger of these, the so-calledA filaments, about 100 A in diameter, are found in the A band arrangedparallel to the myofibril axis. I filaments, about 50 A in diameter, are the1 I<. Bailey, ,4nn. Reports, 1946, 43, 280; D.M. Needham, ibid., 1952, 49, 275.2 S. V. Perry, in “ Comparative Biochemistry,” ed. M. Florkin and H. S. Mason,Academic Press, New York, 1960, Vol. 11, p. 245; idem, Physiol. Rev., 195G, 36, 1;A. F. Huxley, Progr. Rioph?sics Biophys. Cham., 1957, 7, 255344 BIOLOGICAL CHEMISTRY.main components of the I band and continue into the A band to form aninterlocking hexagonal array with the A filaments 394 (Fig. 1).Although isolated myofibrils can shorten by about 75--80~0 of theiroriginal length, in vivo under physiological conditions contraction rarelyexceeds 40% of the resting length. Recent studies of band changes occur-ring during contraction over the physiological range have indicated thatthe A band remains unchanged in length. The whole of the shortening inthis range occurs in the I band, which ultimately disappears when the myo-fibril has contracted by about 35--40y0 of its resting length.To explainthese band changes it has been suggested 3 ~ 5 that the myofibril contracts byvirtue of the I filaments’ moving further into the A band along the spacesbetween the A filaments, thus bringing about the shortening and final dis-appearance of the I band. On relaxation when the myofibril returns to itsoriginal length the reverse process takes place.The problem is to devise a satisfactory physicochemical mechanismwhich will explain how one protein filament system can be drawn deeperinto the other with which it forms an interdigitating hexagonal array. Themechanism must be reversible and be correlated with the chemical changesknown to accompany contraction.All workers agree that the myofibril isbuilt of longitudinal filaments, but although the evidence for the two-fila-ment system described above is excellent for rabbit skeletal muscle,* a num-ber of workers have not been able with the electron-microscope to demon-strate such a two-filament system in certain other striated muscles.6~7 Like-wise in smooth muscle some workers claim that only one type of filament ispresent.* Contraction is characteristic of all types of muscle and, althoughthe ultrastructure of the muscle tissues so far examined consists of fila-mentous protein elements, the electron-microscope evidence available to datedoes not permit us to say that the arrangement of these filaments is identicalin every case.It seems, however, that all muscles contain a commonspecialised biochemical system. The precise proportions and performanceof this system depend on the type of muscle, but it might be supposed thatthe mechanism of contraction at the molecular level is similar in all suchtissues.Chemical natuure of contractile system. The muscle tissues so far examinedcontain three proteins, myosin, tropomyosin, and actin ; these proteins areobtained only from muscle and in the limited studies of isolated myofibrilsthey have been shown to be localised in these contractile elements in thecell and to make up about 85-90y0 of the myofibrillar dry weight. Certainother components are also said to be present in the myofibril and accordinglypresumed to have some special role in contraction.These are usually notwell defined and are present in relatively small amounts2 At presentnothing is known of their function, nor have any special features in their3 J. Hanson and H. E. Huxley, Symposia SOC. Exp. Biol., 1955, 9, 228.4 H. E. Huxley, J . Biophys. Biochem. Cytol., 1957, 3, 631.6 J. L. Farrant and E. H. Mercer, Ex$. Cell Res., 1952, 5, 553.7 A. J. Hodge, J . Biophys. Biochem. Cytol., 1955, 1, 361.8 C. F. Shoenberg, J . Biophys. Biochem. Cytol., 1958,4, 609,A. F. Huxley and R. Niedergerke, Nature, 1954, 173, 971; H. E. Kuxley and J.Hanson, ibid., p. 973PERRY: THE CHEMISTRY OF MUSCLE CONTRACTION AND RELAXATION. 345properties which may play a role in contraction yet been recognised.Pre-cise analysis of the myofibril in terms of its protein components is still notpossible, owing to difficulties of determination, but provisional figures forthe rabbit myofibril are given in the Table.Approximate protein composition of the isolated rabbit myojbril.% Total NMyosin .......................................... 50-55Actin ............................................. 20-25Tropomyosin .................................... 10-15Other components.. ............................ 5-10By selective extraction and use of fluorescent antibodies lo it has beenshown that myosin is localised in the A band, that actin and tropomyosinare certainly present in the I band, and that both are probably present inthe A band.If the model proposed by Huxley et aL4y5 is accepted and ifeach filament system is assumed to have a constant protein composition, it-/,so0 il a23ii Myosin400i - - ISX 7i o p o myosin- 241 Actin dimerS d-/&oooi A Filament - - N O i< * 4 o i-20,000~ T FiIomontFIG. 2. Diagram to illustrate the relative dimensions of filaments of the rabbit myojibriland the molecules of its three main protein components (vertical and horizontai scabsnot identical).follows that myosin is localised in the A filaments and that actin and tropo-myosin are associated with the I filaments. There is some electron-micro-scopic evidence *y7 for interfilamentous material in the A band but nothingis known about its nature.Comparison of the molecular sizes and shapes of the myofibrillar proteinswith the dimensions of the contractile system indicate that they are similar.The molecules are presumably aligned along the axes of the filaments whichcan be at most only a few molecules thick (Fig.2).Of the three major protein components, only actin and myosin havephysicochemical properties which can be readily related to the contractileprocess. The role of tropomyosin is not known but, since the originalisolation from vertebrate muscle of the soluble form, tropomyosin B,ll it9 W. Hasselbach, 2. ges. Naturw., 1953, 86, 449; J. Hanson and H. E. Huxley,10 H. Finck, H. Holtzer, and J. M. Marshall, J . Biophys. Biochem. Cytol,, 1956, 211 K. Baiiey, Biochem. J., 1948, 43, 271.Nature, 1953,172, 530; A.Corsi and S. V. Perry, Biochem. J., 1958, 08, 12.suppl., 175346 BIOLOCICAI. CHBMISTRY.has been shown l2 that certain invertebrate muscles contain an insolubleform known as tropomyosin A, in addition to the soluble type. Tropo-myosin A is found in large amounts in molluscan adductor muscles (up to30% of the total protein) and it has been suggested that it may be respon-sible for the characteristic tonic properties of these m~sc1es.l~ Although thesolubility properties of the tropomyosins are different, the amino-acidanalyses of the two types are fundamentally similar and for this reason theseproteins are known collectively as tropomyosins.Myosin. Myosin has been the subject of considerable physicochemicalinvestigation. Although the most highly purified preparations, L- andcrystalline myosin, used in these studies appear monodisperse on ultra-centrifugation or electrophoresis, evidence is accumulating that they maybe far from satisfactory in this respect.Most work has been carried out onpreparations of rabbit skeletal-muscle myosin which are well known to becontaminated with 5'-adenylic acid deaminase and small amounts of nucleicacid. According to some workers the standard preparations of L-myosincontain 10-15% of impurities,14J5 and very recently Kominz et aZ.16 suc-ceeded in separating from myosin by dialysis against O-lM-sodium carbonatea distinct electrophoretic component of molecular weight 29,000 whichamounts to about 14% of the original myosin. The relation of this proteinto other so-called sub-units of myosin obtained by urea I5,l7 and other treat-ments 16 is not clear.Treatment with urea would not be expected to destroycovalent linkages but the units obtained in this way are smaller than theL- and H-meromyosin produced by tryptic or chymotryptic digestion. Con-trolled heat-denaturation of myosin also facilitates the separation of addi-tional components.l8 It is worth speculating whether L- and H-mero-myosin 19 represent true covalently bound sub-units rather than componentswhich pre-existed in the myosin molecule and whose separation is facilitatedby predigestion.2°*21 Such a consideration may involve a fine distinctionof what constitutes a sub-unit, but recent work on the chromatography ofmyosin on diethylaminoethylcelulose indicates that myosin preparationscontain components of different ATP-ase activity.21 In contrast, H-mero-myosin appeared to be homogeneous with respect to ATP-ase activity.22As yet there has been no clear demonstration of the separation of an ATP-ase of high activity from myosin.When fractionation of the protein occurs12 I<. Bailey, Pubbl. Stax. Zool. Nupoli, 1956. 29, 96; Biochim. Biophys. Ada, 1957,24, 612; D. R. Kominz, F. Saad. J. A. Gladner, and K. Laki, Arch. Biochem. Biophys.,1957, 70, 16; D. R. Kominz, F. Saad, and K. Laki, Conference on the Chemistry ofMuscular Contraction, Japan, 1957, p. 66.13 J. C. Ruegg, Biochim. Biophys. Actu, 1959, 35, 278.14 W. F. H. M. Mommaerts and R. G. Parrish, J , Bid.Chem., 1951, 188, 645.15 T. C. Tsao, Biochim. Biofihys. Ada, 1953, 11, 368.16 D. R. Kominz, W. R. Carroll, E. N. Smith, and E. R. Mitchell, Arch. Biochent.17 A. G. Szent-Gyorgyi and M. Borbiro, Arch. Biochern. Biophys., 1956, 60, 180.18 R. H. Locker, Biochim. Biophys. Acta, 1959, 32, 189.19 A. G. Szent-Gyorgyi, Arch. Biochem. Biophys., 1953, 42, 306; J. Gergely, M. A.20 W. R. Middlebrook, Abs. 4th Internat. Congr. Biochem., Vienna, 1958, p. 84.21 S. V. Perry, Biochenz. J., 1960, 74, 94.25 H. Mueller and S. V. Perry, Biochim. Biophys. Acta, in the press.Biophys., 1959, 79, 191.Goiivea, and D. Karibian, J . BioE. Chem., 1955, 212, 165PERRY: THE CHEMISTRY OF MUSCLE CONTRACTION AND RELAXATION. 347the enzymically active portion represents the major part of the originalprotein.Notwithstanding these reservations about their nature a considerableamount of precise physicochemical study has been applied to myosin pre-parations, Up to recently a molecular weight of 840,000 was accepted23but with clear evidence now available for aggregation in myosin solutionsunder certain conditions 24 this value has become the subject of contro-versy.25 Application of the Archibald approach to sedimentation equili-brium has given values in the region of 420,000,26 and similar values havebeen obtained by the sedimentation-diffusion method at low protein con-centration.Such figures would fit well the present views on the sub-unitcomposition of myosin derived from investigations on the meromyosins.26aPhysicochemical study of N-ethylmaleimide-poisoned myosin in 5~-guan-idine hydrochloride has indicated that the myosin molecule consists of threepolypeptide chains each wound in a tight a-helix, these chains together form-ing a three-stranded rope-likeActiuz.Actin is of particular interest as a myofibrillar component in thatit combines with myosin to form the complex, actomyosin, which possessesthe special property of responding to the addition of ATP under certain ionicconditions by contraction. Russian workers 28 have made the interestingclaims (although recently with some reservation 29) that this property residesin the myosin alone and that the role of actin is to alter and stabilise the pH-dependence of the ATP-ase. Such claims do not fit in with conventionalviews, nor are they supported by the observation of Hayashi et thatactin is an essential requirement for contraction at both pH 7.6 and 9-0.Some further work has been carried out on the nature of the processinvolved in the polymerisation of G(globu1ar)-actin 31 (considered to be adimeric form of particle weight 140,000) to the F(fibrous)-form which isextremely viscous in solution and combines with myosin to form contractileactomyosin.Conventional views would regard this as a linear polymeris-ation of G-actin dimers but polarisation-fluorescence studies 31 suggest thatsome caution must be exercised before this model is adopted. Although thepolarisation-fluorescence studies give a molecular weight for the polymerisedF-actin no greater than that of the dimer, application of the light-scatteringtechnique suggested that it may reach a value of several millions.32From a systematic study of the polymerisation of G-actin, Oosawa and23 H.H. Weber and H. Portzehl, Adv. Protein Chew., 1952, 7, 161.24 A. Holtzer, Arch. Biochem. Biophys., 1956, 64, 507.25 H. H. Weber, Ann. Rev. Biochem., 1957, 26, 667.s6 P. H. Von Hippel, H. K. Schachman, P. Appel, and M. F. Morales, Biochim.Biophys. Acta, 1955, 28, 504; X7. F. H. M. Mommaerts and B. B. Aldrich, ibid., p . 627.865 Cf. S. Lowey and A. Holtzer, Biuckim. Biuphys. Acta, 1959, 34, 470.27 W. W. Kielley and W. F. Harrington, Fed. Proc., 1959, 18, 259.z8 W. A. Kafiani and V. A. Engelhardt, Doklady Akad. Nauk S.S.S.R., 1953, 92, 385.29 W.A, Engelhardt, Conference on the Chemistry of Muscular Contraction, Japan,30 T. Hayashi, R. Rosenbluth, P. Satir, and M. Vozick, Biochim. Biophys. Acta,81 T . C . Tsao, Biochim. Biophys. Acta, 1953, 11, 227.32 J. Gergely and H. Kohler, Conference on the Chemistry of Muscular Contraction,1957, p. 134.1958, 28, 1.Japan, 1957, p. 14348 BIOLOGICAL CHEMISTRY.his co-workers conclude that it can be regarded as a reversible ‘‘ fibrouscondensation.” A critical actin concentration, determined by conditions inthe medium, is required before polymerisation occurs. Above the criticalconcentrations excess of G-actin is converted into the F-form. These in-vestigations suggest that all preparations of F-actin contain both F- andG-actin which are in equilibrium and undergo continuous interconversion.The fact that G-actin preparations contain a small but persistent amountof ATP, which is converted into ADP on polymeri~ation,~~ has stimulatedspeculation on the role of this nucleotide and the G-F transformation inrelation to the contractile process.An implication of these nucleotidechanges during G-F interconversion on the theory of Oosawa et al. is thatdephosphorylation of ATP occurs continuously in F-actin solutions. Theseauthors have claimed that added ATP is slowly dephosphorylated by F-actinsolutions.=As yet there is no direct evidence of participation of the G-F trans-formation in contraction. It would be attractive to consider the nucleotidewhich is bound to the myofibri136 as a built-in acceptor for the phosphatebond energy used in contraction, but such a view is not supported by thefact that the phosphorus of this nucleotide only slowly equilibrates withthat in the nucleotide pool of the muscle cell after injection of inorganicP2PJorthopho~phate.~~ The equilibration is not speeded up significantlyby activity of the muscle.Although the bound nucleotide of the myofibrilis rather inert so far as enzymes are concerned, Strohman 38 has found thatthe nucleotide associated with isolated actin can participate in reactionsinvolving the creatine-phosphokinase system. Pertinent to the problem aresome interesting observations on the localisation of nucleotide metabolismwithin the I band, as revealed by autoradi~graphy.~~Actomyosin. The interaction of actin and myosin and the effects ofATP on it are perhaps the most striking phenomena of muscle biochemistry.Combination of these two proteins causes a marked increase in viscosity..On addition of low concentrations of ATP this is reduced to a value whoselogarithm equals the sum of the logarithms of the viscosities of actin andmyosin, measured separately. Szent-Gyorgyi 40 first explained this effectas a simple dissociation of the complex into actin and myosin, but it hasbeen suggested41 as a result of light-scattering studies that a change inmolecular shape rather than dissociation is responsible for it. This inter-pretation of the light-scattering data is, however, not supported by allinvestigator^,^^ and certainly separation of actin and myosin can be demon-33 S.Askura, K. Hotta, N. Irnai, T. Ooi, and F. Oosawa, Conference on the Chemistryof Muscular Contraction, Japan, 1957, p. 57.34 F. Oosawa, S. Askura, K. Hotta, N. Imai, and T. Ooi, J . Polymer Sci., 1959, 37,323; F. Oosawa, ibid., 1957, 26, 29.85 F. B. Straub and G. Feuer, Biochim. Bio9hys. Acta, 1950, 4, 455.36 S. V. Perry, Biochem. J., 1952, 51, 495.87 A. Martonosi, M. A. Gouvea, and J. Gergely, Fed. PYOC., 1959, IS, 283.38 R. C . Strohman, Biochim. Biophys. Ada, 1959, 32, 436.89 D. K. Hill, J . Physiol., 1959, 145, 132.40 A. Szent-Gyorgyi. Acta Physiol. Scund., 1945, 9, Suppl. No. 25.4 1 J. J. Blum and M. F. Morales, Arch. Biochem. Biophys., 1953, 43, 208.‘2 J. Gergely, J . Biol. Chem., 1956, 220, 917; H.Nuda and K. Maruyama, Biochim.Bio@ys. Acta, 1959, 30, 598PERRY: THE CHEMISTRY OF MUSCLE CONTRACTION AND RELAXATION. 349strated 43 by ultracentrifugation of actomyosin solutions in the presence ofATP.Whereas the interaction of actomyosin and ATP can be more readilyinvestigated by physicochemical methods when it occurs in solution, studyof the reaction ilz v i m , occurring as it does in a heterogeneous system, presentsgreater difficulties. The contracting effect of ATP on actomyosin gels insuitable ionic environment can readily be compared with the behaviour ofmore physiological models such as isolated myofibrils or glycerated musclepreparations.& Any differences which are apparent are probably due todifferences in the degree of orientation of the protein filaments in the twotypes of system.In actomyosin solutions the dissociating action of ATPis reflected in the change of viscosity. Presumably dissociation of the com-ponents on addition of ATP also occurs in the gel but is then followed bycontraction of the system. Any mechanism proposed for contraction in themyofibril system with its precise orientation of filaments and localisation ofprotein must apply also to the randomly oriented actomyosin gel precipitatedfrom a solution of the complex.Protein-protein interactions are not unusual in biological systems but,unlike actomyosin formation, most are non-specific. Myosin thiol (SH)groups are essential for both ATP-ase and actomyosin-forming activity,45and for a number of thiol reagents there is a close correlation between thedegree of inhibition of both properties. With several types of inhibitor thegeneral pattern of behaviour suggested that the same active centres wereessential for actin-combination and enzynie-substrate complex-formationalthough there was some indication that with iodoacetamide the ATP-aseactivity of myosin was less sensitive to this reagent than was the acto-myosin-forming property.Using very high concentrations of the inhibitor,BMny 46 succeeded in treating actomyosin with iodoacetamide and isolatingfrom it a myosin component which had no ATP-ase activity but combinednormally with actin. These results and those obtained with other thiolreagents are taken4' to indicate that different thiol groups on the myosinmolecule are necessary for the interaction with ATP and with actin, but thatthe " pyrophosphate binding " part shares in both activities.Actin alsocontains thiol groups, which are involved in the G-F transformation,48 butthe ability of F-actin to combine with myosin is relatively independent ofthiol reagents.45Adenosine tripkcosphatase. The other aspect of the interaction of myosinwith ATP which seems important for the contractile process is the hydrolysisof the nucleotide :ATP + H,O I_e ADP + H,PO,45 A. Weber, Biochim. Biophys. Acta, 1956, 19, 345; J. Gergely and A. Martonosi,43 H. H. Weber and H. Portzehl, Prop. Biophysics Biophys. Chem., 1954, 4, 60.45 K. Bailey and S. V. Perry, Biochim. Biophys.Acta. 1947, 1, 506.46 M. B&r.r&ny, 4th Internat. Congr. Biochem., Vienna, 1958, Abs., p. 84.47 M. BQrAny and K. B&r&ny, Biochim. Biophys. Acta, 1959, 35, 293.48 G. Feuer, F. Molnar, E. Pettk6, and F. B. Straub, Acta Physiol. Acad. Sci. Hung.,Fed. Proc., 1958, 17, 227.1948, 1, 160; G. ICuschinsky and F. Turba, Biochim. Biophys. Acta, 1961, 6, 426350 BIOLOGICAL CHEMISTRY.The enzyme concerned is perhaps more accurately described as a nucleosidetriphosphatase rather than an ATP-ase as it will also split the triphosphatesof inosine, guanine, uridine, and cytidine at high rates. Inorganic tri-phosphate is hydrolysed slowly. Detailed investigation of the enzyme 2 9 4 4 ~ 4 9has revealed many unusual properties, but despite the mass of experimentalfacts a satisfactory picture of the mechanism of ATP hydrolysis has yet toemerge.Myosin ATP-ase is atypical in its activator requirement in that it isactivated by calcium and not by magnesium although the latter is usuallymore effective with enzymes using ATP as substrate.Magnesium, alone orin the presence of calcium, inhibits the enzyme,50 and in the latter caseantagonism between the ions is a~parent.~l The thiol nature of the enzyme iswell e~tablished,~~ and it is thus surprising that under certain conditions lowconcentrations of specific thiol reagents increase the hydrolysis catalysed bycalcium-activated A T p - a ~ e . ~ ~ Another apparently paradoxical effect isthat a t high ionic strengths low concentrations of ethylenediaminetetra-acetate (EDTA) stimulate the enzyme53 whereas a t low ionic strengthsimilar concentrations inhibit 55 2,4-Dinitrophenol also stimulates thecalcium-activated ATP-ase of myosin 52,56 and it is of interest to comparethis system with the mitochondrial ATP-ase which is likewise sensitiveto dinitr~phenol.~' Earlier ideas 57958 that the mitochondrial ATP-ase mightbe associated with a contractile system regulating volume changes of mito-chondria have recently been discussed further.59Although most of the above observations apply to actin-free myosinpreparations, the presence of actin profoundly affects the enzymic behaviourof the system.In strong contrast to their effect on myosin, both magnesiumand calcium at low ionic strength markedly activate actomyosin ATP-ase;55y60the ions are no longer antagonistic and under certain conditions may besynergic.55 At higher ionic strengths (> 0.15) magnesium-activation dis-appears and the enzymic behaviour more closely corresponds to that of49 D.M. Needham, Adv. E.ttzymoZ., 1952, 13, 151; H. H. Weber and H. Portzehl,Adv. Protein Chem., 1952, 7, 161; A. G. Szent-Gyorgyi, Adv. Enzymol., 1955, 16, 313;K. Bailey, in " The Proteins " (eds. H. Neurath and K. Bailey), Academic Press, NewYork, 1954, Vol. IIB, p. 951; M. F. Morales, J. Botts, J. J. Blum, and T. L. Hill, Plzysiol.lieu., 1955, 35, 475.60 I. Banga and A. Szent-Gyorgyi, Stud. Imt. Med. Chem. Szeged, 1943, 3, 72.51 &I. F. H. M. Monimaerts and K. Seraidarian, J . Gen. PhysioZ., 1947, 30, 401.52 G.D. Greville and D. M. Needham, Biochim. Biophys. Acta, 1955, 16, 284; J. 13.Chappell and S. V. Perry, ibid., p. 285; W. W. Kielley and L. B. Bradley, Fed. Proc.,1955, 14, 235.53 E. T. Friess, Arch. Biochem. Bioflhys., 1954, 51, 17; E. T. Friess, M. F. Rlorales,and W. J. Bowen, ibid., 1954, 53, 311; G. D. Greville and E. Reich, ibid., 1957, 20,440.54 W. J. Bowen and T. D. Kerwin, J . Biol. Chem., 1954, 211, 237.55 S. V. Perry and T. C. Grey, Biochem. J., 1956, 64, 184.66 H, C. Webster, Ph.D. thesis, Cambridge, 1953; S. V. Perry and J. B. Chappell,67 S. V. Perry, Conferences et Rapports, 3rd Internat. Congr. Biochem., Brussels,58 J. B. Chappell, Ph.D. thesis, Cambridge, 1954.59 A. L. Lehninger, Symposium on Molecular Biology (ed.R. E. Zirkle), Univ.60 S. V. Perry, Biochem. J . , 1951, 48, 257.Biochem. J., 1957, 65, 469.1055, p. 365.Chicago Press, Chicago, 1959, p. 122PERRY 1 THE CHEMISTRY OF MUSCLE CONTRACTION AND RELAXATION. 351myosin at a similar ionic strength. These effects apply to actomyosin madeby combination of the separately isolated proteins and to more physiologicalsystems such as isolated myofibrils. Despite the fact that purified myosinis not activated by magnesium this ion is clearly important far functioningof myosin in situ as it is essential for the contraction and relaxation 62 ofisolated myofibrils.With actomyosin systems marked differences occur in the temperaturecoefficient of the calcium- and magnesium-activated action of A T p - a ~ e .~ ~Analysis of the Arrhenius plot for the hydrolysis of ATP by myosin in thepresence of magnesium (a reaction which can hardly be called activated asthe rate is very slow) in the absence and presence of 2,4-dinitrophenol sug-gests that the enzyme-substrate complex with substrates lacking the 6-amino-group may have a conformation which is particularly sensitive tochange a t about 16". Although this phenomenon does not occur with ATPalone, yet if dinitrophenol is added the behaviour now resembles that withITP, suggesting that the dinitrophenol may be strongly attracted to thegroup or groups on the protein which normally bind the amino-group inRTP.63Kinetic analysis of calcium-activated hydrolysis of ATP by L-myosindoes not allow a clear distinction between two possible mechanisms, namely,those wherein (i) the substrate is Ca-ATP and free ATP is inhibitory, or(ii) calcium myosinate is the activated form of the enzyme which splits freeATP, and Ca-ATP is inactive.The latter mechanism has been consideredmore acceptable because of the simpler postulates.M Similar hypotheseshave been put forward 55,65 to explain the experimental findings withmagnesium as the activator for actomyosin ATP-ase. Certain features ofthe latter system, namely, inhibition by EDTA at a concentration one-fiftiethof that of the r n a g n e ~ i u m , ~ ~ * ~ ~ and the relief of inhibition induced by excessof ,4TP by very low concentrations of calcium,55 imply that for the activationby magnesium small amounts of some cation (e.g., Ca2') are required whichcan be selectively removed by the binding action either of EDTA or of ATP.Mreber's recent study 67 of the myofibrillar ATP-ase supports such an ex-planation.This hypothesis can be used as the basis of a plausible theoryof the mechanism of inhibition of ATP-ase which occurs during relaxation(see below).An unusual feature of the hydrolysis of ATP by L-myosin or actomyosinis that the reaction begins with a high initial rate which over the first 10-20sec. may be up to five times as high as the stationary value reached within1-2 m i n u t e ~ . ~ * ~ ~ ~ EDTA eliminates this effect, whereas both the initial61 C. A. Ashley, A. -4rasimavicius, and G. ill. Hass, Exp. Cell. Res., 1956, 10, 1 .62 J.R. Bendall, J. Physiol., 1953, 121, 232.63 H. M. Levy, N. Sharon, and D. E. Koshland, Biochi-m. Biophys. Acta, 1959, 33,,288.64 L. B. Nanninga, Biochim. Biophys. Acta, 1959, 36, 191; Arch. Bioclzem. Biophys.,66 G. Geske, M. Ulbrecht, and H. H. Weber, Arch. Ex$. Pathol. Phaumahol., 1957,1957, 70, 346.230, 301.66 S. V. Perry and T. C. Grey, Biorhenz. J., 1956, 64, 5 ~ .G7 A. Weber, J. Biol. Chem., 1959, 234, 2764.68 ,4. Weber and W. Hasselbach, Biochim. Biophys. Actn, 1964, 15, 237.G9 Y . Tononiura and S. ICitagawa, Hioch~nz. Htophys. d c f a , 1957, 26, 15352 BIOLOGICAL CHEMISTRY.and the stationary phase are enhanced by 9-chloromercuribenzoate and2,4-dinit~-ophenol.~~ If the hydrolysis of ATP is essential for the develop-ment of tension this initial phase of high activity may be of physiologicatsignificance.It would be of great advantage for the myofibril to hydrolyseATP veryrapidly for a few milliseconds during the initial stage of a muscletwitch.Some light has been thrown on the mechanism of ATP hydrolysis bystudying the reaction catalysed by myosin systems in the presence of H,180.When the activator used with actomyosin is calcium, one atom of 180, isintroduced into each molecule of inorganic orthophosphate produced.'* Inthis respect the mechanism of the splitting of ATP is similar to that occurringin phosphokinase systems.71 The value of the ratios of the reactivities ofwater and methanol €or enzymic compared with non-enyzmic hydrolysis ofATP are of a very different order, which suggests that the mechanism ofenzymic hydrolysis involves a specific myosin-water interaction.72 How-ever, if the magnesium-activated hydrolysis by L-myosin, intact lobstermuscle, or actomyosin is studied, appreciable exchange of 1802 betweenH2180 and the inorganic orthophosphate formed is apparent.73 This findingis interpreted as indicating that a phosphorylated intermediate which iscapable of exchanging oxygen with water is formed in these systems.Evi-dence of another kind, namely, the binding of inorganic orthophosphate byH-meromyosin during ATP hydrolysis, has also been taken as evidence fora phosphorylated intermediate.74When 32P was used the evidence found for a phosphorylated intermediatewas somewhat controversial. No exchange between inorganic 32P and ATPor between AD3,P and ATP was observed by Koshland et aZ.,70 whereasWeber 75 reported that with '' Fuadin "-poisoned myosin such an exchangecan be demonstrated.An exchange of 32P between AD32P and ATP whichis stimulated by magnesium has been demonstrated with preparations ofactomyosin extracted as the complex (" natural " actomyosin), and withm y ~ f i b r i l s , ~ ~ ~ ~ ~ This exchange also occurred with myosin extracted selec-tively from whole myofibrils, but not with L-myosin or preparations of" synthetic " actomyosin from purified actin and myosin. The granularATP-ase of muscle actively catalysed this transfer,77 but although thisenzyme contaminates " natural " actomyosin and myofibril preparations itwas considered that the major part of the exchange studied was catalysedby the actomyosin system.So f a r as L-myosin is concerned there is generalagreement that with this protein alone it has not so far been possible todemonstrate any exchange of 32P between AD32P and ATP.70 D. E. Koshland, 2. Budenstein, and A. Kowalsky, J . Biol. Chzem., 1954, 211, 279.71 B. Axelrod, Adv. Enzymol., 1956, 17, 159.73 D. E. Koshland and E. B. Herr, J . Biol. Chem., 1957, 228, 1021.73 H. M. Levy and D. E. Koshland, J . Amer. Chsm. SOC. , 1958, 88, 3164.74 J. Brahms and C. Rzysko, Abs. 4th Internat. Congr. Biochem., Vienna, 1958,75 H. H. Weber, Conferences et Rapports, 3rd Internat. Congr. Biochem., Brussels,76 G. Ulbrecht and M. Ulbrecht, Biochim. Biophys.Ada, 1957, 25, 100.77 G. Ulbrecht, M. Ulbrecht, and H. J. Wustrow, Bioclainz. BiopJays. Acta, 1957, 25,p. 83.1955, p. 356.110PERRY: THE CHEMISTRY OF MUSCLE CONTRACTION AND RELAXATION. 353ATP and contraction. Two important aspects of the interaction of ATPwith the actomyosin system, namely, the mechanical changes and thehydrolysis of ATP, have to be related to the events occurring in vivo. Thereis little doubt that the contraction which can be induced by ATP in acto-myosin systems in vitro is the counterpart of contraction in living muscle.Striking confirmation of this is shown by treating myofibrils with ATPunder controlled conditions : these structures then contract and exhibitthe same band changes as take place in contracting living m u ~ c l e .~ ? ~ ATPis not unique in producing contraction, for other nucleoside triphosphatescan bring about this change; nevertheless the latter are present in musclein concentrations much lower than that of ATP which, if it is not separatedfrom the site of action by some physical barrier, would be expected to be theactive agent purely on mass-action considerations.The question whether ATP is split during contraction, quite apart fromthe fact as to whether it is the contractile agent, is also controversial. As aconsequence of prolonged activity in excess of what the particular muscle isnormally called upon to perform, inorganic orthophosphate accumulates, thelevel of creatine phosphate falls, and subsequently so does the ATP con-centration.2~78 Some investigators 79 have felt that to prove that ATP isthe primary source of energy for the contractile system it is necessary todemonstrate a fall in the level of this substance during the initial stages of asingle twitch, the whole event normally lasting about 100 milliseconds.This is undoubtedly an ideal requirement but the possibility of being ableto demonstrate such a change is doubtful.It is estimatedso that theorthophosphate liberated would represent a very small fraction of the ATPpresent. Hydrolysis of ATP would be occurring under conditions wherethe whole metabolism of the cell is poised ready to rephosphorylate any ADPproduced. Certainly the inability to show an appreciable drop in ATPmust not in itself be taken to exclude the possibility of ATP splitting duringcontraction without ensuring that the enzyme systems concerned withrephosphorylation are ineffective.With in vitro systems considerable correlation between contraction andsplitting occurs.2~23*49 Although there are occasions when correlation isnot satisfactory these may arise because the contractile process is moresensitive to the physical state of the systems than is the enzymic activity.Contraction obviously will not be apparent when the actomyosin is in solu-tion, but such conditions certainly favour ATP-ase activity.A more seriousobjection would be the demonstration that contraction can be induced byATP in a model system without its being hydrolysed. Evidence of thiskind has yet to be obtained.It may be concluded that studies with in vitro systems, in which it ispossible to evaluate accurately the ATP level and the rate of its hydrolysis,there is reasonably good correlation between contraction and tension de-veloped on the one hand and simultaneous hydrolysis of ATP on the other.78 W.F. H. M. Mommaerts, “ Muscular Contraction,” Interscience Publ. Inc., New79 E.g., A. V. Hill, Nature, 1949, 163, 320.80 W. F. H. M. Mommaerts, Amer. J . Physiol., 1955, 182, 585.York, 1950.REP.-VOL. LVI 354 RIOLOGICAL CHEMISTRY.Some exceptions exist, but in the opinion of the Reporter these are notserious objections to this hypo thesis.With intact living muscle, measurement of chemical change during asingle twitch presents considerable difficulties.They arise because of therelatively small change which may be expected, its extremely short duration,and problem of fixing the muscle at a given time (measured in milliseconds)to prevent further chemical change. Chemical reactions in the restingmuscle approach a steady state and it is impossible to predict the preciseeffect of a single twitch on this condition. It is conceivable that thenucleoside-polyphosphate level is relatively insensitive to a low level ofactivity. Within the last 10-15 years a number of attempts have beenmade to tackle this problem and, except in two recent investigations,s0S8lsome evidence for ATP breakdown has been presented. It is, however,difficult to compare all the investigations because the conditions and degreeof activity to which the muscle was subjected were somewhat variable.Theexperiments of Fleckenstein et aZ.B1 and of Mornmaerts:O involving directanalysis of the phosphate compounds of muscle most likely to be hydrolysedduring contraction, have shown that various skeletal muscles of the turtleand frog rectus abdominis can undergo a single twitch without significantchange being apparent in the ATP, ADP, or creatine-phosphate levels. Onthe other hand, indirect determination of nucleotide changes during contrac-tile activity in intact living muscle by spectrophotometric means has pro-vided evidence for a small increase of ADP level after a single twitch. Thisincrease is estimated to be 2% of what would be expected on the assumptionthat the energy required for the work done by the muscle during contractionwas derived from ATP.82Even if breakdown of ATP in living muscle is masked by the efficiency ofrephosphorylation systems, activity should give rise to an increased turnoverin the phosphate atoms of creatine phosphate and of the muscle nucleotides.As yet, however, tracer studies have failed to reveal a transfer of phosphatefrom phosphocreatine to ATP either during drug-induced contracture offrog muscles 83 or during tetanic contracture of cat ga~trocnernius.~~ Evenmore difficult to reconcile with the hypotheses that ATP is dephosphorylatedduring contraction is the further finding g5 that after injection of inorganic[32P]orthophosphate stimulation more than 10,000 times in one hour didnot cause a significant change from the normal resting condition in thedistribution of 32P between creatine phosphate, the terminal phosphate ofADP, and the terminal and middle phosphate of ATP. In neither restingnor stimulated muscle were any of the organic phosphates in equilibriumwith the total orthophosphate.Difficulties arise in interpreting suchstudies, for much of the injected isotope is in the extracellular spaces andnot in metabolic contact with the nucleotides within the cell. This heavilylabelled extracellular phosphate will form part of the orthophosphate81 A. Fleckenstein, J. Janke, R. E. Davies, and H. A. Krebs, Natztre, 1954,174, 1081.82 B. Chance and C. M. Connelly, Nature, 1957, 179, 1235.S3 A. Fleckenstein, J. Janke, R. E.Davies, and W. Richter, Arch. Ex*. Pnthol.8s J. Sacks, Nrrfure, 1959, 183, 825.Phavmakol., 1956, 228, 596.G. J. Dixon and J. Sacks, Amer. J . Physiol., 1958, 193, 129PERRY: THE CHEMISTRY 01; MWSCLE CONTRACTION AND RELAXATION. 355fraction separated from the muscle. As perfusion of the muscle to removeextracellular phosphate also causes changes in the intracellular organo-phosphates some workers 86 consider that it is not possible to determine theabsolute turnover rate of any organophosphate compound which has in-organic phosphate as its immediate source of phosphorus. Equilibrationbetween injected inorganic p2Plphosphate and muscle nucleotides is rela-tively slow compared with that in other tissues; 87 even so it has been con-cluded 86 that the speed at which the terminal phosphate of ATP is replacedin resting muscle is such as to preclude any attempt to measure differencesin rates in working muscle.Nevertheless, if the results obtained by Sacksand other workers are valid, then it is clear that dephosphorylation of ATPis not the direct source of energy for muscle activity.Some effort has been directed towards a search for alternative sourcesof energy, but as yet such a compound has not been discovered. It is un-likely that carnosine phosphate is present in muscle in appreciable a m o u n t ~ , ~ ~ aand the recently discovered phosphorylated guanidino-compounds 87b foundin some invertebrates are of the phosphagen type rather than substanceswhich directly supply energy for contraction.Relaxation.-When the stimulus reaches the muscle the electrical changesoccurring a t the membrane initiate in some way the chemical changes whichtake place at the myofibril and contraction results.These changes continuefor only a limited period after a single stimulus; in a matter of millisecondsthe tension in the contractile unit drops. There is good evidence 88 thatrelaxation is a passive process and extension to the resting length follows asa consequence of the load on the muscle when the contracting force nolonger acts. It seems likely that in resting muscle the enzyme systemswhich play a part in contraction are in steady-state equilibrium. Theelectrical changes occurring a t the membrane as a consequence of the nerveimpulse reaching the cell initiate some slight but significant change, possiblyionic in nature, which completely alters the enzymic balance at the myofibril.A view which has much to commend it is that the myofibrillar ATP-ase isgreatly activated above the low basic level characteristic of the resting myo-fibril, and that shortening then ensues.The initiating electrical changes a tthe membrane are of short duration and, once they are over, the tendencyis for the system to return to the condition characteristic of resting muscle;tension falls and the myofibril extends to its resting length.In earlier work with actomyosin model systems and glycerated fibres 2$ 23949conditions were devised in which previously contracted actomyosin systemscould be made to relax. Usually this occurred in the presence of ATPand various enzyme inhibitors such as EDTA, I ‘ Salyrgan,” etc., whichbrought the ATP-ase activity of the system to a low level.In some casesrelaxation could be obtained in the absence of ATP but in the presence of86 E.g., A. H. Ennor and H. Rosenberg, Biochem. J., 1954, 56, 308.87 K. K. Tsuboi, Arch. Biochem. Biophys., 1959, 83, 445.87b N. V. Thoai, J. Roche, Y . Robin, and N. V. Thiem, Compl. read. SOC. Biol., 1953,147, 1241; N. V. Thoai and Y. Robin, Biochim. Biophys. Acla, 1954, 14, 76; G. E.Hobson and K. R. Rees, Biochem. J . , 1955, 61, 549.88 A. V. Hill, Proc. K0-v. Soc.. 1949, R, 136, 420.D. F. Cain, A. M. Delluva, and R. E. Davies, Nature, 1958, 182, 720356 BIOLOGICAL CHEMISTRY.pyrophosphate and triphosphate 23 which dissociate actomyosin but are notthemselves split. From such studies with model systems the concept hasemerged of the dual role of ATP, i.e., it is able to act, depending on theconditions, as both a contracting and a relaxing agent.When ATP-aseactivity is high, ATP acts as a contracting agent; when it is low, ATPplasticises the system, presumably by breaking the link between actin andmyosin, the tension drops, and the actomyosin fibre relaxes.An important step towards more physiological systems was made whenit was shown that addition of ATP to crude muscle-cell fragments SB and topartly washed glycerated fibres produced an elongation rather than a con-traction. If the systems were more highly purified, addition of ATP pro-duced the well-known contraction.The implication was that these crudepreparations contained some factors which modified the effect of ATP onthe actomyosin system. This factor (known variously as the Marsh factor,Marsh-Bendall factor, or relaxing factor) is considered to be effective inliving resting muscle and prevents contraction from taking place, althoughthe ATP concentration irt sit% in the absence of such a factor would causethe myofibrils to contract.Ideally, the relaxing factor should be studied in systems in which itseffect on the tension of the contracted system can be measured directly.This involves using fibres with which, unless they are extremely thin, itseems impossible to maintain the ATP Concentration constant throughoutthe fibre. ATP is diffusing into the fibre and at the same time being brokendown by the ATP-ase activity of the actomyosin component.For thisreason certain phosphorylase preparations such as creatine phosphokinase 91and myokinase 92 have been claimed to possess relaxing-factor activity. Itseems possible that these phosphokinases may act by keeping the ATP at asufficiently high concentration throughout the system.* Pertinent to thisquestion are the findings of Japanese workersg3 that in addition to thephosphokinase system another muscle fraction, which appeared to be relatedto the granular fraction of muscle sarcoplasm, was necessary. Portzehl 94also found the relaxing factor in the granular fraction which could be sedi-mented completely from sarcoplasm by high-speed centrifugation.Apartfrom causing relaxation of fibre models, relaxing-factor preparations inhibitin a parallel manner the ATP-ase activity of myofibrils 94 and of actomyo~in,~~and the synzeresis of actomyosin gels.g6 Whenever relaxation occursin vitro the evidence so far indicates that ATP-ase activity is low, whichsuggests that inhibition of the enzymic activity brings about the loss in89 B. B. Marsh, Nature, 1951,167, 1065; idem., Biochim. Biophys. Acla, 1952, 9, 247.90 E. Bozler, Amer. J . Physiol., 1951, 167, 276.91 M. C. Goodall and A. G. Szent-Gyorgyi, Nature, 1953, 172, 84; L. Lorand, itid.,p. 1181; E. Bozler, J . Gen. Physiol., 1954, 37, 63.92 J. R. Bendall, Proc. Roy. Soc., 1954, B, 142, 409.93 H.Kumagai, S. Ebashi, and F. Takeda, Nature, 1955, 176, 166; S. Ebashi,F. Takeda, M. Otsuka, and H. Kumagaia, Symposia on Enzyme Chem., Japan, 1956,94 H. Portzehl, Biochim. Biophys. Acfa, 1957, 26, 373.D. J. Baird and S. V. Perry, Biochem. J., in the press.H. Mueller, Biochim. Biophys. Ada, in the press.11, 11.* At high ATP concentrations the magnesium-activated ATP-ase is inhibitedPERRY: THE CHEMISTRY OF MUSCLE CONTRACTION AND IIELAXAllON. 357tension and resulting relaxation of the system.97 Portzehl 98 concluded thatif sufficiently thin fibres were used there was no requirement for a phospho-kinase system as well as the granular preparation of the relaxing factor,either to bring about relaxation in fibres or to inhibit the ATP-ase activityof myofibrils.More recently, however, Molnar and Lorand 99 have providedfurther evidence for the potentiating action of pyruvic phosphokinase withrelaxing-factor preparations. Nevertheless, other workers 1oo-102 considerthe phosphokinase function in the system to be non-specific, but theirexperiments suggest that there is a requirement for a dialysable cofactor inthe system.The precise centrifugal force required to sediment granules (or possiblyreticular material) with relaxing-factor activity from muscle homogenatesdepends somewhat on the species and type of muscle used.95 In the liter-ature the activity is often considered to be associated with the microsomefraction, but studies on the distribution in rabbit skeletal muscle render thisless certain.The bulk of the activity in rabbit skeletal muscle is sedimentedbelow 20,000 9, and the fraction rich in oxidative activity is also the mostconcentrated with respect to relaxing factor. The factor is less easily sedi-mented from pigeon breast-muscle homogenates than from similar prepar-ations of rabbit skeletal muscle, and there is a sharper distinction betweenit and the fraction rich in oxidative activity. In any case muscle is poor ina granular fraction rich'in nucleic acid and corresponding to the conventionalmicrosome fraction of other ti~sues.10~The factor preparations are lipoprotein in nature and possess ATP-aseactivity 9591019104 which is not derived from myosin and is now known to beassociated with the granular components of s a r c ~ p l a s m .~ ~ ~ Both ATP-aseand relaxing-factor activity are destroyed by phospholipase, but undercertain conditions the relaxing-factor activity can be preferentially de-stroyed.lo6 As yet there has been no report of the preparation of therelaxing factor in a soluble form. A significant finding is that relaxing-factor activity of these preparations is readily abolished by low concen-trations of c a l c i ~ m . ~ ~ ? ~ ~ ~ When assayed in the presence of 5-millimolarsodium oxalate, preparations are effective in extremely low concentrations,which suggests that the function of oxalate is to remove some substance(perhaps calcium) present in the preparation and inhibiting their a~tivity.~5For example, with preparations of granules from rabbit muscle 50% in-hibition of the myofibrillar ATP-ase associated with 1 mole of myosinPhysiol., 1953, 37, 53.n7 J.X. Bendall, J. Physiol., 1953, 121, 232; E. Bozler and J . '1'. Prince, J . Gen.ga H. Portzehl, Biochim. Biophys. .4cta, 1957, 24, 474.gg J. Molnar and L. Lorand, Nafuve, 1959, 183, 1032.loo F. N. Briggs, G. Kaldor, and J. Gergely, Biochim. Biophys. Acta, 1959, 34, 211.lol J. Gergely, G. Kaldor, and F. N. Briggs, Biochim. Biophys. Acta, 1959, 84, 218.lo2 G. Kaldor, J. Gergely, and F. N. Briggs, Biochim. Biophys. Acla, 1959, 84, 224.lo3 S. V. Perry and M. Zydowo, Biochem. J., 1959, '72, 682.lo8 L. Lorand, J. Molnar, and C. Moos, Conference on the Chemistry of Muscularlo5 W. W. Kielley and 0. Meyerhof, J . Biol. Chewz., 1948, 176, 591; S.V. Perry,lo6 S, Ebashi, Arch. Biochem. Biophys., 1958, 76, 410.lo' E. Bozler, Amer. J . Physiol., 1952, 168, 760.Contraction, Japan, 1957, p. 85.Biochim. Biophys. A d a , 1952, 8, 499358 BIOLOGICAL CHEMISTRY.(molecular weight 420,000) is obtained when 20 kg. of total microsomalprotein is present, of which the relaxing factor probably represents only asmall part .95In the absence of oxalate much larger amounts of granules are requiredto bring about inhibition and under such conditions activity falls off onageing at 0°.g5 Such a loss in activity is evident only when assays arecarried out in the absence of oxalate. These results are compatible with theslow liberation, in relaxing-factor preparations, of an ion such as calciumwhich inactivates the preparation unless chelating agents are present.Pyrophosphate behaves in a similar manner to oxalate and to the cofactorreported by Kaldor et aL102If speculation is justified at this stage the following mechanism appearsplausible.In view of the inhibition of the magnesium-activated myo-fibrillar ATP-ase with low concentration of EDTA or higher concentrationsof ATP,55366 it is possible that traces of calcium or of a similar cation areessential in some way for the enzyme.67 The relaxing factor may inhibitmyofibrillar ATP-ase in a similar way to these substances by binding tracesof a cation (perhaps calcium) essential for enzymic activity. Further sup-port for this hypothesis would come from the demonstration that relaxingfactor preparations have a strong affinity for calcium.It is striking that theinsoluble granular preparations can exert their influence on the splittingof ATP which occurs on another insoluble system, the myofibril. Thissuggests that the relaxing factor brings about inhibition, not by direct inter-action at the active centres of the enzyme, but rather by changing thecommon soluble environment of the two systems to one which is unfavour-able for ATP hydrolysis by the magnesium-activated enzyme. The actionsof oxalate and the soluble cofactor lo0-lo2 are sufficiently similar to suggestthat their function is to bind the cation (perhaps calcium) present, whichwould otherwise inactivate the relaxing factor. As a physiological counter-part of oxalate the cofactor could have this function in resting muscle, butcould lose it momentarily when muscle is stimulated so that the myofibrilcan hydrolyse ATP at a high rate and hence contract.Conclusions.-The theories which have been proposed (see Perry for areview) to explain the mechanism of contraction usually involve two mainassumptions about the ultrastructure of the contractile unit. These arethat it consists of a single-filament system which shortens by folding in someway or that it is a two-filament system in which one type of filament movesalong the The latter mechanism has much to commend it; but,although the evidence is good for its occurreiice in rabbit skeletal muscle,conclusive evidence for a similar system in other types of muscle has yet tobe produced.It seems reasonable to suppose that the mechanism of con-traction a t this level of organisation is common to all muscle tissues. Bothactin and myosin (and possibly other myofibrillar proteins) appear to beactive participants in contraction, and the two-filament mechanism utilisesthe interaction between the two proteins in a convincing way to explain bothcontraction and relaxation. ATP is usually given a role in modern theorieslo8 €1. H. Weber, ‘‘ The Motility of Muscle and Cells,” Harvard Univ. Press, Cam-bridge, Mass., 1958FOWDEN NEW AMINO-ACIDS FROM PLANTS. 359but not all workers consider its dephosphorylation to occur simultaneouslywith contraction. Present views on enzymic activity during relaxationstrengthen the case for this hypothesis; and, in addition, ATP has uniqueproperties as a relaxing agent compared with other nucleoside triphos-phates.log Much of the work at the growing points of muscle biochemistrymay seem imprecise by purely physicochemical standards, but this is aconsequence of the complexity of the systems studied and the problems in-volved.It can be said, however, that progress in this field towards a mole-cular biology-the integration of the chemical events and the ultrastructureof a complex biological system-is certainly as well advanced here as any-where in biochemistry, s. v. P.5. NEW AMINO-ACIDS FROM PLANTSDURING the last decade about fifty amino- or imino-acids have been newlyidentified as components of higher plants. About twenty more have beenrecognised either as constituents of micro-organisms or as fragments ofantibiotics excreted by them.As a group, these newly discovered acidshave no striking chemical or physiological properties, and their rapid dis-covery is the result of the application of paper and ion-exchange chromato-graphy to the examination of plant extracts. Brief accounts of the chemistryof some of the newly recognised acids have appeared in earlier Reports,l andtheir biological importance has been considered in other reviews.2 Only afew of the acids are distributed widely in plants; the majority are found onlyin occasional plant species. Their distribution follows no rules. Certaincompounds are characteristic of particular plant families, e.g., azetidine-2-carboxylic acid for the Liliaceae.The distribution of many others is hap-hazard, i.e., they are accumulated in high concentration by only a fewspecies that are botanically quite unrelated. The random distribution hasfavoured the idea that they are unimportant and perhaps ‘ I accidental ”products of metabolism ; however, this concept may prove unacceptablewhen inore plant species have been examined and when more crucial inform-ation is available concerning their metabolic relationships. This Reportwill consider recent contributions to our knowledge of the chemistry andbiochemistry of each of the main types of acid.Dicarboxylic Acids-The only new dicarboxylic amino-acid identified asa constituent of higher plants since the last Reports 1 is a-m-carboxyphenyl-gly~ine,~ which was isolated from the acid fractions of an extract of Irisbulbs.Comparison with synthetic material prepared from m-carboxy-benzaldehyde by a Strecker reaction proved its identity.Two acids have been identified as products of microbial metabolism.@-Methylaspartic acid was formed as an intermediate in the reversibleanaerobic conversion of glutamate into mesaconate and ammonia by Clostri-.IDS W. Hasselbach, Biochim. Biophys. Acta, 1956, 20, 355.Ann. Reports, 1955, 52, 271; 1957, 54, 276.F. C. Steward and J. I<. Pollard, Ann. Rev. Plant. Physiol., 1957,8,65; L. Fowden,C. J. Morris, J. 17. Thompson, S. Asen, and F. Irreverre, J . Anzep. Chenz. Soc.,BioE. Rev., 1958, 33, 393.1969, 81, 6069360 BIOLOGICAL CHEMISTRY.diwn tetanomorphum extract^.^ The acid was provisionally assigned theL-tho-configuration.After growth of Streptomyces ~imosus,~ an actino-mycete better known as the producer of the antibiotic terramycin, substantialquantities (1-2 g./l.) of (+)-Ed-diaminosuccinic acid were isolated from thefermentation medium. New syntheses of meso- and racemic ad-diamino-succinic acids have been published.6Synthetic y-hydroxyglutamic acid has been resolved into its four opticalisomers.' One pair of diastereoisomers formed a lactone very easily andthis provided a basis for their separation. After conversion into chloro-acetyl derivatives, the resolution of each pair of diastereoisomers into D-and L-forms was undertaken by using a hog renal L-acylase preparation.The specific rotations of all the isomers were given.By comparison ofspecific rotations in water and 5~-hydrochloric acid, the natural acidisolated from Hemerocallis was shown to be L-allohydroxyglutamic acid.*A new synthesis for my-methyleneglutamic acid has been published andthe racemic mixture has been res~lved.~ The synthetic method of Hellmannand Lingens lo has been shortened by substituting ethyl a-iodomethyl-acrylate (prepared from a-iodomethylacrylic acid 11) for the analogousbromo-compound. The modified synthesis was used to prepare DL-Y-methylene[a-14C]glutamic acid by condensation of the iodo-intermediatewith diethyl acet amido [ a-14C] malonate. *The metabolism of these glutamic acid derivatives has been studiedrecently.When plants assimilate 14C02 photosynthetically, glutamic acidnormally becomes labelled rapidly, but when 14C02 was supplied to leavesof Phlox decussata,l2 Adiantum j5edatum,12 Lilium regale,13 or tulip l4 littleincorporation of the carbon-14 into y-hydroxyglutamic acid, y-methylene-glutamic acid, or y-hydroxy-y-methylglutamic acid was observed. Experi-ments with [carboxy-14C]pyruvate have provided more definite informationabout the biosynthetic pathways leading to the acids. After labelledpyruvate had been supplied to the fern, Asplepziurn septentrionale, the specificactivity of y-hydroxy-y-methylglutamic acid was higher than that of otherfree amino-acids; l5 a similar result was obtained for y-methyleneglutamicacid and y-methyleneglutamine in ground-nut leaves.16 These facts suggestthat the basic carbon skeleton may be produced by condensation of twomolecules of pyruvate (De Jong17 observed an analogous slow chemical4 H.A. Barker, R. D. Smyth, E. J. Wawszkiewicz, M. N. Lee, and R. M. TVilson,Arch. Biochem. Biophys., 1958, 78, 468.6 H. McKennis and A. S. Yard, J . Org. Chem., 1958, 23, 980.7 L. Benoiton, M. Winitz, S. M, Birnbaum, and J. P. Greenstein, J . Amev. Chew.8 L. Fowden, unpublished result.Y. Nakagawa and T. Kaneko, J . Chem. Soc. Jafian, 1957,78, 232; T. Kaneko andlo H. Hellmann and F. Lingens, Chem. Ber., 1956, 89, 77.l1 K. N. Welch, J., 1930, 257.12 G. E. Hunt, Plant Physiol., 1958, 33, suppl., xii.14 L. Fowden and F. C. Steward, Ann. Bot. N.S., 1957, 21, 69.15 P.Linko and A. I. Virtanen, Acla Chem. Scand., 1958, 12, 68.16 L. Fowden and J. A.- Webb, Ann. Bot. N.S., 1958, 22, 73.17 A. W. K. De Jong, Rec. Trav. chim., 1900, 19, 259.F. A. Hochstein, J . Oyg, Chern., 1959, 24, 679.Soc., 1957, 79, 6192.Y . Nakagawa, ibid., p. 1216.M. E. Wickson and G. H. N. Towers, Canad. J . Biochern. Physiol., 1956, 34, 502kOWDEN: NEW AMINO-ACIDS FKOM PLANTS. 361condensation of pyruvic acid in the presence of gaseous hydrogen chloride).Certainly one molecule of pyruvate must enter y-hydroxy-y-methylglutamicacid and y-methyleneglutamic acid intact and not after conversion intoacetyl-CoA. y-Hydroxyglutamic acid may be formed by condensation ofpyruvate with glycine or glyoxylic acid, but negative results were obtainedwhen labelled substrates were supplied to P.decussata.15When y-methylene[a-14C]glutamic acid was supplied to leaves of pea,tulip, or peanut, carbon-14 was incorporated fairly quickly into a varietyof amino-acids, sugars, and organic acids.8 The relative constancies of theconcentrations of y-methylene-glutamic acid and -&tarnine present inexcised tulip leaves during storage l4 must then be maintained by dynamicequilibria and not by the metabolic inertness of the two substances.The degradation of y-hydroxyglutamic acid in plants has not beeninvestigated, but when y-hydr~xy[a-~~C]glutamic acid was supplied to rats,substantial amounts of the original 14C activity appeared rapidly in glutamicand aspartic acid.18 Enzymes present in extracts of animal livers converthydroxyproline into y-hydroxyglutamic acid via its y-semialdehyde in amanner analogous to that involved in the interconversion of proline andglutamic acid.It is possible that the proline-glutamic acid enzymes actin a non-specific manner by catalysing the conversion of the hydroxy-corn pound^.^^y-Hydroxy- and y-methylene-glutamic acid readily donate their amino-group to a-oxoglutarate in the presence of extracts of Phlox 2o and peanut 21leaves respectively. Ellis 22 has shown that both transamination reactionsare catalysed by a purified aspartate-glutamate transaminase prepared fromcauliflower buds. Since this plant material is not known to contain eithery-hydroxy- or y-methylene-glutamic acid, the reactions catalysed by extractsof Phlox and peanut may be the result of non-specific enzyme action.Both y-methylglutamic and a-aminoadipic acid occur in certain higherplants. The structural relation existing between the two acids is the sameas that between (3-methylaspartic and glutamic acids, and therefore inter-conversion of the six-carbon acids may be shown ultimately to proceed by amechanism similar to that observed for the five-carbon acids in C.tetano-rnorph~m.~Imino-acids.-Three new derivatives of proline have been isolated fromseaweeds. 3-Carboxymethyl-4-isopropenylproline occurs in two forms(L-a-kainic acid, and L-a-allokainic acid; I) in Digenea simplex.* InL-a-kainic acid, the 2- and 3-substituents are trans to one another and thoseat C(,) and C(4) are cis; both configurations are trans in L-a-allokainic acid.The substances have been ~ynthesised.~~ 3-Carboxymethyl-4-(2-carboxy-Is L.Benoiton and L. P. Bouthillier, Canad. J . Biochem. Physiol., 1956, 34, 661.l9 E. Adams, R. Friedman, and A. Goldstone, Biochim. Biophys. Acta, 1958, 30, 212.2o A. I. Virtanen and P. K. Hietala, Acta Chem. Scand., 1955, 9, 549.21 L. Fowden and J. Done, Nature, 1953, 171, 1068.22 R. J. Ellis, personal communication.23 S. Murakami, T. Takemoto, and 2. Shimuzu, J . Pharm. SOC. Japan, 1953,73, 1026.24 Y. Ueno, K. Tanaka, J. Ueyanagi, H. Nawa, Y . Sanno, 5%. Honjo, R. Nakamori,T. Sugawa, M. Uchibayashi, K. Osugi, and S. Tatsuoka, Proc. Japan Acad., 1957, 33,53; K. Tanaka, &I. Miyamoto, M.Honjo, H. Morimoto, T. Sugawa, and M. Uchi-bayashi, ibid., p. 47362 BIOLOGICAL CHEMISTRY.l-methylhexa-l,3-dienyl)proline (domoic acid; 11) was obtained fromChondria a ~ r n a t a . ~ ~ The three substances have useful anthelminthic pro-perties.Hydroxypipecolic acids have received considerable at tention in the pasttwo years. 4-Hydroxypipecolic acid was isolated first from Acacia penta-dena.26 A hydroxypipecolic acid, isolated from Armeria mnritirna,27 wasprovisionally assigned the 3-hydroxy-configuration. Comparisons of thetwo materials showed them to be very similar, if not identical. Subsequentisolations from Acacia willardia and Lysilornn bahamense 28 have providedadditional support for the 4-hydroxy-structure. More recently a hydroxy-pipecolic acid was isolated from A.ecelsn, and evidence supporting a trans-4-hydroxy-configuration was 0btained.2~ 4-Oxopipecolic acid has beenshown to occur in the antibiotic, staphylomycin; 30 on catalytic hydrogen-ation ~-cis-4-hydroxypipecolic acid was formed. cis-3-Hydroxypipecolicacid has been prepared31 and shown to be separable from the imino-acidof Armeria by paper chromatography.8 The position regarding the naturaloccurrence of the 3-hydroxy-compound is now uncertain. However,3-hydroxypicolinic acid occurs in the antibiotics etamycin 32 and staphyl-om ycin .31The trans-configuration assigned to 4-hydroxypipecolic acid brings thiscompound into line with natural 5-hydroxypipecolic acid isolated fromdates; the trans-configuration of hydroxyl and carboxyl groups in the lattercompound is established unequi~ocally.~s A new method of obtaining thediastereoisomeric mixture of (&)-5-hydroxyallo(cis)- and (&)-5-hydroxp-pipecolic acid from kojic acid is available.= (&)-5-Hydroxypipecolic acidwas prepared essentially free from the allo-isomer by reduction of 5-OXO-piperidine-2-carboxylic acid (obtained from glutamic acid) with sodiumb~rohydride.~~ The related dehydro-derivative, baikiain (1,2,3,6-tetra-hydropyridine-2-carboxylic acid), has also been synthesised by a newmethod.After hydrolysis of the benzoyl group from cis-5-benzamido-l-bromopent-3-yne-l-carboxylate, base-catalysed elimination of hydrogenbromide and ring closure gave b a i k i a i ~ ~ ~ Unexpected difficulties were25 K.Daigo, J . Pharm. SOC, Japan, 1959, 79, 353, 356.26 A. I. Virtanen and S. Kari, Acta Chem. Scand., 1955, 9, 170.27 L. Fowden, Biochem. J., 1958, 70, 629.28 A. I. Virtanen and R. Gmelin, A d a Chem. Scand., 1959, 13, 1244.$9 J. W. Clark-Lewis and P. I . Mortimer, Nature, 1959, 184, 1234.80 H. Vanderhaeghe and G. Parmentier, Symposium on Peptide Antibiotics, 17th,31 H. Plieninger and S. Leonhauser, Chem. Ber., 1959, 92, 1579.8% J. C. Sheehan, H. G. Zachau, and W. B. Lawson, J . Amer. Chem. Soc., 1958, 80,53 B. Witkap and C. M. Foltz, J . .4nzer. Chem. SOC., 1957, 79, 192.H. C. Beyerman, Rec. Tvav. chim., 1958, 77, 249.95 H. C. Beyerman and P. Boekee, Rec. Trav. chtm., 1959, 78, 648.36 N. A. Uobson and R. A. Raphael, J., 1958, 3642.Congr.Pure Appl. Chem., 1059, p. 56.3349FOWDEN : NEW AMINO-ACIDS FROM PLANTS. 363found when attempts were made to cause the cyclisation of other 1,5-sub-stituted cis-pent-3-ene-l-carboxylic acids.The biogenetic pathways leading to the heterocyclic ring systems ofproline and pipecolic acid have common features. The possible pathwaysby which pipecolic acid may be formed are illustrated in the annexed chart.The conversion of lysjne into pipecolic acid has been demonstrated byisotopic tracer experiments with higher plants, Neurospora, and intact rats.Reaction (4) is probablyCH ZH2C' '7H2IH02C ,CH-C02HNH2CHZHO-H2C )CH.C02H\H2t/\CHz 2,NH2the primary one occurring i n Neurospora 37 andReactions: ( I ) +2H. (2) -2H.(3) Transaminase or amine oxidase. (4) a-Amino-acidoxidase. (SC) Spontaneous cyclisation. (5) +2H. (6) +2H.rats.38 Loss of the c-amino-group of lysine [reaction (3)] is catalysed by aplant amine oxidase from peas; 39 transamination could also yield cc-amino-adipic 8-semialdehyde. The reduction of l-amino-5-hydroxyhexanoic acid(hexahomoserine) by extracts of N . cyassa [reaction (2)] also produced thealdehyde; DPNH acted as a better hydrogen donor than TPNH.40 a-Amino-8-hydroxyvaleric acid is reduced to glutamic y-semialdehyde undersimilar conditions. As yet, reaction (1) does not appear to have beendemonstrated with certainty, but the analogous conversion of glutamic acidinto its y-semialdehyde is well established.Enzymic reduction of 3,4,5,6-tetrahydropyridine-2-carboxylate intopipecolic acid [reaction (6)] has been observed.41 Extracts of rat organs, ofN .crassa, and of seedlings of pea and green gram (Phaseolus radiatus) aregood sources of the dehydrogenase, which can use either DPNH or TPNH.Reaction (5) is not proved but the corresponding reduction of A4-pyrroline-2-carboxylic acid has been demonstrated for N . crassa 42 and crude extractsof rat organs.41 Ultimately, reaction (5) may be shown to occur in higher37 R. S. Schweet,.J. T. Holden, and P. H. Lowy, J . Biol. Chem., 1954, 211, 517.s8 M. Rothstein and L. L. Miller, J . Amer. Chem. SOL, 1954, 76, 1459.3B P. J. G. Mann and W. R. Smithies, Biochem. J., 1955, 61, 89; A. J. Clark and40 T. Yura and H. J. Vogel, J , Biol. Chem., 1959, 234, 339.41 A.Meister, A. N. Radhakrishnan, and S. D. Buckley, J . Biol. Chem., 1957, 229,T. Yura and 11. J. Vogel, J . Bid. Chenz., 1959, 234, 335.P. J. G. Mann, ibid., 1959, 71, 596.789364 BIOLOGICAL CHEMISTRY.plants since the necessary substrate, 2,3,4,5-tetrahydropyridine-Z-carboxylicacid, is presumably produced by the action of plant amine oxidase.The degradation of pipecolic acid has not been studied extensively.After [l*C]- and pH]-pipecolic acid had been supplied to phyllodes of AcaciahomaZophyUa, radioactivity appeared quickly in various amino-acids, sugars,and organic acids. 4-Hydroxypipecolic acid and a compound presumed tobe 5-amino-l-hydroxyhexanoic acid were amongst the first compounds tobecome labelled in the tritium experiments.* a-Aminoadipic acid and lysinelater became labelled and so some of the above reactions are probablyreversible.Two general mechanisms can be postulated for the biogenesis of thehydroxy-imino-acids.The first requires that the heterocyclic ring is formedonly after the hydroxy-group has been introduced into an appropriate open-chain amino-acid. y-Hydroxyglutamic acid and 6-hydroxylysine, both ofwhich occur naturally, could then yield hydroxyproline and 5-hydroxy-pipecolic acid by reactions analogous to those involved in the synthesis ofproline and pipecolic acid. However, there is no evidence that the con-version of hydroxyproline into y-hydroxyglutamate (see above 19) can bereversed, even in animal tissues. The metabolism of 6-hydroxylysine hasnot been investigated.According to the second mechanism the hydroxyl group is introducedafter the formation of the heterocyclic ring.This mechanism for hydroxy-proline synthesis has now been shown to occur in animal and plant tissues.Hydroxyproline is a rare component of plant proteins but occurs as a char-acteristic component of the protein present in abnormally growing cells,e.g., in tissue cultures and plant tumours. By using cultures of carrot root,it has been shown that hydroxyproline is produced by hydroxylation ofproline only after the latter imino-acid has been incorporated into the cellprotein.& Free hydroxyproline causes inhibition of growth by suppressingthe incorporation of proline into protein.The experiments in which labelled pipecolic acid was supplied to Acacia 8(see above) indicated that 4-hydroxypipecolic acid was formed directly fromthe parent imino-acid.When [14C]lysine was supplied, labelled pipecolicacid was formed very rapidly; the carbon-14 was intraduced subsequentlyinto the 4-hydroxy-compound. No active hydroxylysine could be detected.Degradation of hydroxyproline has been reported for two systems. TheD-amino-acid oxidase of sheep kidney converted either D-hydroxyproline orD-allohydroxyprohe into pyrrole-2-carboxylic acid ; presumably 4-hydr-oxy-A1-pyrroline-2-carboxylate is an intermediate which then spontaneouslyloses water.& A strain of Pse.udmmas, isolated from soil, also producedpyrrole-2-carboxylic acid from L-hydroxyproline or ~-allahydroxyproline,which were interconverted by an epirnerase.44 An alternative degradativepathway was present in Pseudamoaas; the A1-pyrroline ring could open and,after loss of ammonia and subseqaent oxidation, a-oxoglutarate wasformed, as shown opposite.45* J.K. Pollard and F. C. Steward, J. Ex$. Bot., 1959, 10, 17.44 A. N. Radhakrishnan and A. Meister, J . Biol. Chem., 1957, 226, 569.45 E. Adams, J . Biol. Chem., 1869, 234, 2073FOWDEN NEW AMINO-ACIDS FROM PLANTS. 365The biosynthetic mechanism leading to azetidine-2-carboxylic acid isstill not clear. Aspartic acid, ay-diaminobutyric acid, and homoserine arepossible precursors, but when these acids were supplied as 14C-labelled com-pounds to leaves of Co.tzvaZlaria. majalis (lily-of-the-valley), which containedconsiderable amounts of azetidine-2-carboxylic acid, only slight incorpor-ation of radioactivity into the imino-acid was observed.46 Diaminobutyrategave a labelled compound that could be catalytically hydrogenated to yieldazetidine-2-carboxylic acid ; by analogy with proline and pipecolic acidmetabolism, this compound may be Al-azetidine-2- or -4-carboxylic acid.Later work showed that roots, and not leaves, may be the main site forsynthesis of azetidine-2-carboxylic acid in these plants.47Only slight degradation of [14C]azetidine-2-carboxylic occurred inC.maj'alis leaves in 48 hours.46 In contrast, a soil yeast degraded azetidine-2-carboxylic acid rapidly with initial formation of y-amino-or-hydroxy-butyric acid; this was subsequently converted into p-alanine, possibly wiuy-amino-a-o~obutyrate.~8Diamino- and Basic Acids.-Recently several additions have been madeto this group of naturally occurring acids.ap-Diaminopropionic and ory-diaminobutyric acid, both previously recognised only as components ofantibiotics, have now been identified as products from higher plants. Theformer was isolated from seeds of Mimosa palmeri; 49 the latter was presentin traces in the rhizome of Polygonaturn multijfor~rn.~~ New chemical syn-theses are available for diaminopropionic acid 51 and for diaminobutyricacid.52Derivatives of ap-diaminopropionic acid and the homologous acids,ornithine and lysine, have been isolated. L-( -)-a-Amino-13-ureidopropionicacid (albizziine, the lower analogue of citrulline) occurs in large quantitiesin seeds of various Albixxia species and those of other Mimosa~eae.~~953Albizziine was converted into diaminopropionic acid by treatment withboiling 48% hydrobromic acid.The isomeric L- (3-amino-a-ureidopropionicacid has been ~ynthesised.~~8-Acetylornithine occurs in ferns, grasses, and many members of the*6 L. Fowden and M. Byrant, Biochem. J., 1959, 71, 210.47 L. Fowden, Biochem. J., 1959, 71, 643.48 H. Vinson and L. Fowden, unpublished result.*9 R. Gmelin, G. Straws, and G. Hasenmaier, 2. physiol. Chem., 1959, 814, 28.50 L. Fowden and M. Bryant, Biochem. J., 1958, 70, 626.51 H. Hellmann and G. Haas, Cheun. Ber., 1957, 90, 1357.53 G. Talbot, R. Gaudry, and L. Berlinguet, Canad.J . Chem., 1958, 36, 593; M.Fraenkel, Y. Knobler, and T. Sheradsky, Bull. Res. Council Israel, Sect. A. Chem., 1958,7, 173; M. Zaoral, Chem. Listy, 1958, 52, 2338.53 A. Kjaer, P. 0. Larsen, and R. Gmelin, Experientia, 1959, 15, 253366 RIoLocIc.u, CHEMISTRY.Fumariaceae.u The flagellar proteins of the bacterium, SaZmoneZZa typhi-murium, yield E-N-methyl-lysine on hydrolysis ; 55 this amino-acid was notdetected in the protein component of the remainder of the bacterial cell.y-Guanidinobutyric acid is present in a variety of plant tissues; 56 from1 to 20 pg./g. fresh weight occur in the fruit of many species. The acid maybe formed by transamidination known to occur in animal tissues; can-avanine, arginine, or guanidinoacetic acid can donate their amidine groupto ornithine, canaline, glycine, or hydr~xylamine.~~ The formation ofy-guanidinobutyrate from y-aminobutyrate and arginine is catalysed byextracts of rat or pig kidney.58 Glycine was better than, and P-alanineinferior to, y-aminobutyrate as an acceptor of the transferred amidine group.A deguanidinase, differing in its properties from arginase, has been found invarious fish livers ; this enzyme hydrolytically splits y-guanidinobutyric acid,yielding urea and y-amin~butyrate.~~Canavanine, an important constituent of jack-bean seeds (Canaualiaensiformis), has now been shown to occur in a wide variety of other legu-minous seedsm A new type of cleavage of canavanine to homoserine andhydroxyguanidine occurs in a pseudomonad.61 Kalyankar et aL61 alsosummarise the other known degradative reactions of canavanine that yielda variety of products including O-ureidohomoserine, guanidine, canaline, andCC-oxo- 8-guanidinox ybut yric acid.Another guanidino-compound, y-hydroxyarginine, has been isolated fromthe marine animal, Polycheira rufescens (the sea-cucumber) .62 Alkalinehydrolysis produced y-hydroxyornithine, an amino-acid whose naturaloccurrence would not be unexpected, and which bears the same structuralrelationship to hydroxyproline as 8-hydroxylysine does to 5-hydroxy-pipecolic acid.Acids derived from A1anine.-Amino-acids containing a benzenoid orheterocyclic ring attached to the @-carbon atom of an alanine residue featureas protein components (e.g., phenylalanine, tyrosine, tryptophan, andhistidine).Recently several additional alanine derivatives have beenisolated from higher plants; they appear to occur only as free amino-acids.2,4-Dihydroxy-6-methylphenylalanine (111) has been obtained fromseeds of Agrostemma githago (corn cockle).63 The acid was synthesised bythe annexed route.An amino-acid (stizolobic acid) containing a y-pyrone nucleus has beenisolated from Stizolobium hassjoo,64 a shrubby legume; the plant also con-tains 3,4-dihydroxyphenylalanine. The new acid, p- (3-carboxy-y-pyron-54 A. I. Virtanen and P. Linko, Acta Chem. Scand., 1955,9,531; L. Fowden, Nature,1958, 182, 406; R. H. S. Manske, Canad. J . Res., 1937, 15, B, 86; G. Reuter, Flora,1957,145, 326.55 R. P. Ambler and M.W. Rees, Nature, 1959, 184, 56.5g F, Irreverre, R. L. Evans, A. R. Hayden, and R. Silber, Nature, 1957, 180, 704.1: 3 . 1. Pisano, C. Mitoma, and S. Udenfriend, qdzkre, 1997, 180, 1125.59 R. Baret and M. Mourgue, Compt. rend. SOC. R i d , 1957, 358, Flj.61.6o E. A. Bell, Biochem. J., 1958, 70, 617.61 G. D. Kalyankar, M. Ikawa, and E. E. Snell, J . Biol. Chem., 1958, 239, 1175:6* Y . Fugita, Bull. Chem. SOC. Japan, 1959, 82, 43.9.63 G. Schneider, Biochem. Z . , 1958, 330, 42B:64 S. Hattori and A. TGpmrnipe, Nptu-fe, 1959, 183, 1136., B. Walker, J . Biol. Chem., 1957, 224, 57FOWDIJN : NEW AhiTh’O--4.\C‘IDS 1;EOkl PLAKTS. 3675-y1)alanine (IV), is only a minor component and was isolated from theexudate flowing from cut sterns of seedlings. Ozonolysis of stizolobic acidgive a mixture of aspartic, oxalic, and formic acid.eo -0H=C: IN-CFhHO G C H ; .CH. I CO2HMe N% (111 1Another heterocyclic ring (the pyrazole nucleus) is present in p-pyrazol-1 -ylalanine (V), an amino-acid isolated from seeds of Citn.zZZus vulgaris(watermelon). This is the first report of a naturally occurring pyrazolederivative. Smaller quantities of the acid are present in seeds of relatedcucurbitaceous plants. The acid was synthesised by refluxing s i l ~ e rpyrazole with ethyl a-amino-p-chloropropionate. The infrared spectrum ofthe synthetic material was identical with that of the natural acid after itsra~emisation.~~ The compound isolated from watermelon juice, and pro-visionally identified as p-imidazol-l-ylalanine,66 was probably the abovepyrazole derivative.A new hetero-cyclic alanine derivative recently isolated fromA .willardia has been identified as L-p-uracil-3-1 ylalanine (willardiine; VI).66n The new acid is anisomer of mimosine (leucanol), first isolated fromNH2 Mimosa pudica.66b Willardiine and mimosine mayboth be regarded as derivatives of ap-diamino-propionic acid.The biosynthesis and biodegradation of these alanine derivatives havenot been investigated.Hydroxy-amino-Acids.-Certain hydroxy-amino-acids have been de-scribed under earlier headings. Two further hydroxy-acids have beenobtained from higher plants. y-Hydroxyvaline was isolated from Kalanchoedaigremoiztiana; its presence in other Kalanchoe species could not be con-but by paper chromatography small amounts of its lactone wereThe genus Albixxia is a rich source of newer amino-acids.N‘COHc\ ,N*CH2-CH’Co2H ICH(VI)c3 I;.I:. No6 and L. Fonden, Naizwe, 1959, 184, ~ . 4 . 69.66 S. Shinano and T. Kaya, J . Agric. Chem. SOC. Japaw, 195i, 31, 759.66n R. Gmdlin, 2. phvsiol. CItem., 1969, 316, 164.G6b J: Renz, %. phyiiol. Chem., 1936, 244, 153;67 J . K. I’ollard. E, Sondheimer, and F. C . Steward, Na144re, 1958, 182. 1360,R. -4dams and J. I,. Johnson, J .Anirv. Chem. Sor., 1949, 71, 705368 BIOLOGICAL CHEMISTRY.identified in extracts of K . daigremontiana. y-Hydroxyvaline was syn-thesised by catalytic hydrogenation of P-methyl-a-oxo-y-butyrolactone togive its a-hydroxy-lactone ; this was converted into the a-chloro-derivativeby treatment with thionyl chloride in pyridine, and y-hydroxyvaline wasthen obtained after treatment with concentrated ammonia.The stereo-isomeric composition of the product was not investigated.The second compound isolated was O-acetylhomoserine ; 68 like homo-serine, it was isolated from pea plants. The function and metabolism of thetwo hydroxy-acids have not been studied.p-Hydroxyleucine has been identified as a component of a peptide-typeantibiotic produced by a strain of PaeciZ0myces.6~Aminobutyric Acids.--Aminobutyric acid is plresent in the soluble-nitrogen fraction of almost all plants. It may be formed from glutamic acidby the action of glutamic decarboxylase, an enzyme widely distributed inplants.y-Aminobutyrate is produced by other pathways in micro-organisms , e.g. , by transamination from succinic semialdehyde in Pseudo-monas juorescens 70 and ToruZopsis ~ t i l i s , ~ ~ by hydrolysis of the lactam ringof 2-pyrrolidone in P. aer~ginosa,~~ and by oxidation of pyrrolidine orputrescine via y-aminobutyraldehyde in P. f l ~ o r e s c e n s . ~ ~ It is doubtfulwhether these mechanisms are of any importance for the synthesis of y-amino-butyric acid in higher plants, but putrescine, and possibly succinic semi-aldehyde, have been identified in some plants.The suggestion has been made that y-aminobutyric acid may be re-carboxylated to yield glutamic acid. Comparison of the metabolism ofy-amino[I4C] butyrate and [14C]glutamine in tissue culture of carrot roothas lent support to this idea,73 but only slight reversal of normal glutamic-decarboxylase activity has been dem~nstrated.~~ However, when theenzymic re-carboxylation of y-aminobutyric acid was studied in the presenceof an anion-exchange resin, larger proportions of glutamic acid were pro-duced (the glutamate formed was absorbed by the resin and rendered un-available to the enzyme).75 A similar situation may be operative in livingtissues where cellular organisation may ensure that newly formed mole-cules of glutamic acid are removed immediately from the site of enzymeaction.It is more probable that y-aminobutyric acid is converted into succinicsemialdehyde by a transamination and that the aldehyde is then oxidised tosuccinic acid.Many micro-organisms, including yeast, bacteria, andunicellular algae, contain transaminases capable of catalysing the reactionbetween y-aminobutyrate and a-oxoglutarate. The properties of theenzyme present in P. $uorescens have received detailed study.7o The sameN. Grobbelaar and F. C. Steward, Nature, 1958, 182, 1358.69 G. W. Kenner and R. C. Sheppard, Nature, 1958, 181, 48.70 E. M. Scott and W. B. Jakoby, J . Bid. Chew., 1959, 234, 932; W. B. Jakoby7 1 R. Pietruszko and L. Fowden, unpublished result.73 F. C. Steward, R. G. S. Bidwell, and E. W. Yemm, J. Exp. Bot., 1958, 9, 11.74 R. Koppelman, S. Mandeles, and M. E. Hanke, J . Biol, Chew., 1958, 233, 73.75 J. K. Pollard, personal communication,and J. Fredericks, ibid., p. 2145.F.F. No6 and W. J. Nickerson, J. Bad., 1958, '45, 674FOWDEN: NEW AMLNO-ACIDS FROM PLANTS. 369authors studied the properties of succinic semialdehyde dehydrogenase,isolated from this organism.76The metabolism of y-amin~[carboxy-~~C]butyric acid has been studied inthe organism, Bacillus ~ ~ m i l u s , ~ 7 which produces and excretes large amountsof glutamic acid into the culture medium. If glutamic acid was formedprimarily by direct carboxylation from this specifically labelled y-amino-butyrate, then the o-carboxy-group would have carried the heaviest label-ling. But over 99% of the activity present in the glutamate was associatedwith the a-carboxy-group, which is in accord with its formation from thecarbon skeleton of y-aminobutyric acid via the reactions of the tricarboxylicacid cycle.When the organism was grown in the presence of some of theorganic acid intermediates of this cycle, the glutamate excreted had, asexpected, a lower specific activity.y-Amino[l4C]butyric acid was metabolised quite readily when suppliedto pea leaves; aspartic acid, alanine, and glutamic acid became radio-active in succession,78 as would be expected if the carbon skeleton wasmetabolised via the tricarboxylic acid cycle. This sequence of labellingsuggests that the initial degradation of y-aminobutyrate involved a trans-amination; direct carboxylation would give glutamic acid as the primarylabelled compound.Convincing demonstrations of a y-aminobutyrate-transaminase in higherplant tissues are, however, rare, A transaminase, catalysing a reactionbetween y-aminobutyrate and a-oxoglutarate, was found in extracts ofnodulated pea but further experiments have shown that the enzymeactivity is confined to the nodules.78 Since the microbial symbiont, Rhizo-biz~um, contains an active y-aminobutyrate-transaminase, the transaminationobtained u7ith nodulated roots may have been caused entirely by the infectiveorganism.A similar transaminase has been reported in a potato tuberextract,80 but this observation is of doutbful value because microbial con-tamination was not excluded during the long, 24 hour, reaction period. Withmitochondria from peanut seedlings, definite transamination betweeny-aminobutyrate and either a-oxoglutarate or pyruvate has been obtainedin 3 hours.78 Pyruvate was approximately five times more efficient thana-oxoglutarate as an amino-group acceptor.Oxidative degradation of y-aminobutyric acid has not been demonstratedas yet in plants, but oxidative mechanisms operate in brain tissue and leadto the formation of y-amino+- and -cc-hydroxybutyric acid.81 Eventuallythese mechanisms may be shown to be widely distributed.By the pro-duction of these p- and a-hydroxy-acids, y-aminobutyrate metabolism usesintermediates common to degradation of hydroxyproline 4* and azetidine-2-carboxylic acid,48 respectively. Mutant strains of Escherichia coli alsoproduce y-amino-cc-hydroxybutyric acid by decarboxylation of y-hydroxy-76 W. B. Jaboby and E. M. Scott, J . Biol. Chem., 1959, 284, 937.77 T.Tsunoda and I. Shiio, J . Biochem. (Japan), 1959, 4$, 1011, 1227.78 R. 0. D. Dixon and L. Fowden, unpublished result.79 J. I<. Miettinen and A. I. Virtanen, Acta Chem. Scand., 1953, 7, 1243.80 T. Suzuki, A. Maekawa, T. Hasegawa, M. Ito, H. Honda, T. Nagano, S. Saito,and Y . Sahashi, Bull. Agric. Chem. SOC. Japan, 1958, 22, 39.81 K. Inui Med. J . Osaka Univ., 1959, 11, 681; S. Sao, ibid., 1957, 7, 833370 I~TOI~OGICAL CHEMISTRY.glutamic acid,82 and the 8-hydroxy-acid similarly from allo-p-hydroxy-glu tamate .83Onemethod used y-aminobutyrate which, after N-acetylation, was brominatedon the cc-carbon atom. The bromine atom was replaced by a hydroxylgroup and acid-catalysed de-acetylation yielded the required a ~ i d . 8 ~ It hasalso been prepared by treating ay-diaminobutyric acid with nitrous acid.48a- and p-Aminoisobutyric acid are both known as plant products.Thep-amino-compound was isolated recently from the bulbs of I ~ i s tingitana; 8ssynthetic and racemised natural material gave identical infrared spectra.A new synthesis of the acid from glycine involving Wolff rearrangement ofthe diazoethyl ketones has been described.g6 a-Aminoisobutyric acid is acomponent of the antibiotic produced by Paecilomy~es.~~ The metabolismof these acids has not been investigated.Acids derived from Cysteke.-The isolation of (+)-S-methyl-L-cysteinesulphoxide from cabbages and turnipss7 has been followed by that of itsprobable biological precursor (-)-S-methyl-L-cysteine. The latter acid wasisolated from seeds of Phaseolus vulgaris (kidney beans),88 and occurs as ametabolite of N.crassa. It can act as a sole source of nutrient sulphur forthis organism.8g A transmethiolase, isolated from yeast, has been partlypurified and shown to catalyse the formation of S-methyl-L-cysteine fromL-serine and methanethi01,~O No other amino-acid tested could substitutefor L-serine as an acceptor of the MeS group. Ethanethiol reacted a t about60% of the rate observed for methanethiol. y-L-Glutamyl-S-rnethyl-L-cysteine has also been obtained from kidney-bean seeds; 91 and the corre-sponding sulphoxide occurs in the Lima bean.g2The seeds of Albixzia julibrissin contain S-2-carboxyethyl-~-cysieine inamounts equal to 0.3% of their dry weight.93 The acid has been synthesisedfrom L-cysteine and p-bromopropionic acid.94 A crude enzyme preparationfrom A .lophantha seeds causes hydrolysis of S-2-carboxyethyl-~-cysteine toammonia, pyruvic acid, and p-mercaptopropionic acid.93 S-2-Carboxy-propyl-L-cysteine has been identified tentatively as a component of ,4.willadia.66aCycloalliin (5-methyl-l,4-thiazan-3-carboxylic acid l-oxide) (VII), iso-lated recently from onion bulbs (Allium ~ e p a ) , ~ ~ may be regarded as a cyclisedy-Amino-a-hydroxybutyric acid has been synthesised only recently.82 A. I. Virtanen and P. K. Hietala, Acta Chem. Scand., 1955, 9, 549.83 W. W. Umbreit and P. Heneage, J . Biol. Chem., 1953, 201, 15.84 A. Mori, J . Biochem. (Japan), 1959, 46, 59.85 S. Asen, J. F. Thompson, C . J. Morris, and F. Irreverre, J . Bid. Chem., 1959, 234,86 K. Balenovie, I. JambreSiC, and I. Ranogajec, Croat Chem. Acta, 1957, 29, 87.87 R. L. M. Synge and J. C. Wood, Biochem. J . , 1956, 64, 252; C. J. Morris andJ. F. Thompson, C. J. Morris, and R. M. Zacharius, Nature, 1956, 178, 593.89 J. B. Ragland and J. L. Liverman, Arch. Biochem. Biofihys., 1956, 65, 574.E. C . Wolff, S. Black, and P. F. Downey, J. Amer. Chem. SOC., 1956, 78, 5958.91 R. M. Zacharius, C. J . Morris, and J. I;. Thompson, Arch. Riochenz. BioFhp., 1959,343.J. F. Thompson, J . Sac, Chem. Ind., 1955, 951.80, 199.H. Rinderknecht, J . Soc. Chem. Iltd., 1957, 1354.R. Gmelin, G . Strauss, and G. Hasenmaier, 2. Nafwforsch., 1958, 13b, 252%94 A. Schoberl and A. Wagner, Z . physiol. Chew., 1956, 304, 97.95 A. I . Virtanen and E. J . Matikkala, Acin Chfwz. Scnizd., 1959, 13, 623Some properties of recently isolated amino- and imino-Amino- and imino-acid M. p.*m-Carboxyphenylglycine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . .(+)-aa'-Diaminosuccinic acid . . . . . . . . . . . . . . . . . . . . . . . . . . . 240-290 dL-y-Allohydroxyglutamic acid , . . . . . . . . . . . . , , . . . . . . . . . 187 dL-y-Methyleneglutamic acid , . .. .... . .. . . . .. . . . . . . . . . . . 195-197 dL-a-Kainic acid .. ... . ... . ... . . . . . . . ... .. . ... . ..... ...... . . .. 251 dL-a-Allokainic acid ... . ... . . . . . . . ... .. . ... ..... . .... .. .. .. 237 dDomoic acid ................................................ 217 dL-5-Hydroxypipecolic acid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 d~-trans-4-Hydroxypipecolic acid . , . . . . . . . . , . , . . . . . . . . 294 dL-ay-Diaminobutyric acid dihydrochloride . . . . . . . . . 197-198L-ap-Diaminopropionic acid monohydrochloride . . . 237 dalbizziine . . . . . . . . . . . . . . . . , , . . . . . . . . . . . . . . . . . . . . . . , . . . . , . . . . . 2 17 d6-Acetylornithine . . . . . . . . . . . , . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . 266 dy-Hydroxyarginine monohydrochloride . . . . . . . . . . . . 190-191 d/3-(2,4-Dihydroxy-6-methylphenyl)alanine . . . . . . . . . 252 dStizolobic acid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231-233j?-Pyrazol-l-ylalanine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236-238 d~-fl-Uracil-3-ylalanine . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . 204-205 dy-Hydroxyvaline .......................................... 228/3-Aminoisobutyric acid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183S-Methyl-L-cysteine . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , 169-1 70S-Methyl-L-cysteine sulphoxide . . . . . . . . . . . . . . . . . . . . . . . . 220 dy-Glutamyl-S-methyl-L-cysteine . . . . . . . . . . . . . . . . . . . . . 165-173 dS-2-Carboxyethyl-~-cysteine . . . . . . , . . . . . . . . . . . . . . . . . . , . 2 18 d215" dL-threo-fl-Methylaspartic acid . . , . . . . . . . . . . . .. . . . , , .. . . . , -O-Acetylhomoserine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -Cycloalliin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -[.ID inH2O- 10"- 12- 15+8 - 109.6-23.1- 13.1---+ 13.3 -- 67 + 13.1 ---- 73+ 10- 13- 26 + 125- 21----fa]^ in HC1(normality inparentheses)Approx.in phenol-H'ZO3-13.3" (5N)3-59 ( 1 . 5 ~ )-1-3 (5N)+14 (3N) ---9 (N) -0*0.82(+)0.33(-)0.31(-)0.43(0-0.86(+)0-0.80(-)0.56(0.16(-)0*0.69(0.82(-)0.710.62(0*0.35(@* bl. p.'s associated with decomposition are indicated as d. t Phenol-H,O mixtures normally contain 75--80% (w/w) of phenol.$ Composition of solvent mixture used varies slightly between different laboratories,+ or - in parentheses indicatesabsence of NH, respectively.Colour symbols: Y, yellow; Br, brown; R, red; V, violet; B, blue; P, purple; G, green372 BIOLOGICAL CHEMISTRY.form of alliin, S-allyl-L-cysteine sulphoxide, obtained earlier from garlic(AZZium s a t i ~ u m ) . ~ ~ Cycloalliin was synthesised by the following stages : 95L-cysteine, S-allyl-L-cysteine, S-2-bromopropyl-L-cysteine, 5-methyl-1,4-thiazan-3-carboxylic acid (VIII) (by ring closure involving elimination ofhydrogen bromide in pyridine) , cycloalliin [formed from (VIII) by treatmentwith hydrogen peroxide]. Cycloalliin was decomposed by refluxing it withG~-hydrochloric acid, and oxidised and reduced products were obtained(see formulae).SO3H SO3HI II IQ I 1H2C/Sx7H2 + CHI + CHIN H26N-IH2C s, CH 2I INMe-CH C HsC02H Me-HC, ,CH.C02H HC' Me*HC ,N,CH.C02HH H NH2(VII1) g - M e t h y l - Cysteictaurine a c i dThe biosynthesis of cycloalliin may proceed by hydration of alliin toyield S-2-hydroxypropyl-~-cysteine sulphoxide ; then elimination of waterfrom the hydroxyl and a-amino-groups may cause ring closure.Summary of Properties.-The Table annexed lists some of the physicalproperties of amino- and imino-acids isolated recently. Unfortunately,complete data are frequently not published, sometimes owing to lack ofmaterial. To the biologist, RF values are often more useful than m. p.'s anddata are given here for two commonly used solvents. Differences in theexperimental procedures used in various laboratories cause minor variationsin these values, and, in order to minimise these variations, RF values obtainedwith butan-l-ol-acetic acid-water mixtures are based upon that of alanine(Ralsnine). The colour produced from each amino-acid spot on paperchromatograms treated with ninhydrin is somewhat variable; as far aspossible those given in the Table are quoted from the original literature.L. F.L. FOWDEN.J. GLOVER.T. W. GOODWIN.R. L. HARTLES.S, V. PERRY.96 A. Stoll and E. Seebeck, Experientia, 1947, 3, 114